Am I the Only One That Feels This Way?
Do you suffer from early season regret? Or is it just me?
I’m a value-based drafter. I don’t go into the draft with a plan. I have my values and I sit back, watch the draft unfold, and let the opportunities come to me. I’m also an accountant. I love Excel.
In other words, I’m boring.
The old guys. The boring guys. The “cool” players from two years ago or maybe even ten years ago… That’s the type of player that usually ends up on my teams. Certainly not the fun new toys. I’m told it’s the smart way to play the game. I’m stock-piling value. Increasing my chances of winning.
And I hate it.
I’m also pushing 35. So maybe I’m experiencing early onset mid-life crisis. But I’m second guessing my way of doing business. You might have seen a little bit of the second guessing here. And then I went through a 10 team AL-only draft and it really sent me into a tail spin. Take a look at what’s causing me to question my purpose in life:
| My Boring Team | Price Paid | My Value | That Guy’s Sexy Team | Price Paid | My Value |
| Masahiro Tanaka | $17.00 | $18.11 | Marcus Stroman | $23.00 | $14.82 |
| Michael Pineda | $10.00 | $14.09 | Luis Severino | $14.00 | $7.71 |
| Ian Kennedy | $7.00 | $9.38 | Carlos Rodon | $10.00 | $8.03 |
| Erasmo Ramirez | $2.00 | $1.13 | Jose Berrios | $5.00 | $0.00 |
| Tyler Skaggs | $2.00 | $0.00 | Blake Snell | $3.00 | $0.00 |
| Jimmy Rollins | $1.00 | $1.00 | A.J. Reed | $3.00 | $0.00 |
| Nori Aoki | $1.00 | $11.45 | Jackie Bradley Jr. | $3.00 | $8.12 |
| Adam Lind | $2.00 | $6.20 | C.J. Cron | $3.00 | $5.82 |
| Josh Hamilton | $1.00 | $0.00 | Jurickson Profar | $1.00 | $0.00 |
| Total | $43.00 | $61.36 | Total | $65.00 | $44.50 |
Look at all those shiny toys to play with on the other guy’s team! Well, maybe one day he’ll get to play with them. But still! Look at all of that mouth-watering and potentially game-changing production on the sexy team.
And outside of Pineda (who is a little bit exotic?), look at that boring group of minivan driving Dads with side parts on my team.
I’m comforted by the excess value my calculations show. But owning that other team looks like so much more fun.
Is There a Weakness in My Game?
The real reason for this anxiety is that I have been flat out crushed in the standings by the sexy team the last two seasons. Players like Danny Salazar, Kole Calhoun, Josh Donaldson, George Springer, Corey Kluber, Garrett Richards, and Mookie Betts have slaughtered my Prince Fielder, Robinson Cano, Evan Longoria, Brett Gardner, Billy Butler, and Howie Kendrick.
Hindsight makes it easy to look at the boring names of the last few years and laugh. And to still get excited about the popular or flashy picks. But at the time, I was able to secure the boring guys at discounts and those trendy picks were going at above market prices. The problem is that to some extent the overpays paid off.
I can’t help but feel that there are weaknesses in the steady and boring approach. Or maybe it’s not a weakness. Maybe I’m adhering to it too strictly and leaving myself open to the knockout punch? Maybe I’m drafting too much for the projected value and not considering upside enough?
Or maybe I’m overreacting. Maybe I’ve just had bad luck the last two seasons on my big investments and I shouldn’t change my process. I keep hearing from the DFS community that I shouldn’t necessarily change my process due to bad outcomes. But it sure gets tough.
My Second Issue
Tell me if this story sounds familiar. I do the diligent thing. Draft solid players that are definitely ownable (above replacement level) and would be hard to drop… While some others walk away from the draft with names that are well below replacement level and should not have been drafted.
This is a clear advantage for me, right? Then why am I so envious of those that have obvious players to drop in the early weeks of the season?
A few of my leagues drafted weeks ago. Since then, I’ve felt handcuffed with guys I’m hesitant to drop and had to watch names like Trevor Story, Juan Nicasio, Socrates Brito, Jeremy Jeffress, Shane Greene, Keon Broxton, and Tyler Goeddel get picked up. This week I’ve seen Luke Gregerson (argh, I should have seen that coming!), Tony Zych, Jake Lamb, and Sam Dyson get snagged. Are many of them likely to stick and be long-term contributors this season? Probably not. But a few of those players are “going to pop”. Big. And I won’t have them.
Should I?
Stop Whining and Do Something About It
This turns out to be great advice to my seven year old and a mantra I can use to improve at fantasy sports.
Before I come up with a plan of action, let’s reflect on why this is a problem and what could be going wrong. I can think of three things that could be happening here.
The first is that I’m not being aggressive enough on upside. How much “upside” does one need to win a league? I need to investigate this further. These players getting snagged early in the season are all upside. They’re essentially free and they appear to have been bestowed with significant amounts of playing time. The only cost they have is the opportunity cost of having to drop a player. Which brings me to my next point.
The second possibility is that I’m overvaluing my own players. I’m sure this is part of it. I want to be right about the players I drafted. This is probably a cognitive bias that I’m susceptible to (maybe we all are?). And it’s easy to dismiss these hot pickups as the flavor of the week that won’t pan out. I have a sense that I overvalue my own players when working through trades with others. So it only makes sense that I’d do that when considering adds and drops. I need to work on a way to be more objective about things and remove this bias.
Or third, maybe I’m not quick enough or accurate enough in adjusting my projections and values of these players. Related to this, I have a hard time distinguishing between the immediate benefit of adding a guy like a Socrates Brito and a longer term speculative play like a Tony Zych or Sam Dyson.
Counterpoint!
Is there really a problem here? Why am I so convinced that my process is flawed? Because I had two bad seasons? Maybe things are working fine and it’s been bad luck. Could my process be so flawed that spending $40 to get $60 of value is an inferior strategy to spending $60 to get $40 of value?
That seems pretty stupid. Maybe I need to tweak things a little bit. BUT MY PROCESS CAN’T BE THAT FLAWED.
When I read this great article by Renee Miller (see tweet below, it’s the “Don’t derail your draft” article), I felt quite a bit better. It seems like my current strategy is insulated from the effects of group think and the availability heuristic (read that article if you don’t know what these are!), and maybe I need to stick to my guns.
Two #FantasyBaseball strategy articles @ESPNFantasy
Don’t derail your draft https://t.co/GIiq2l5uqE
&
Cognitive Biashttps://t.co/HddFyE6bU3— Renee Miller (@reneemiller01) March 26, 2016
I’ve Calmed Down. Now I’ll Act.
Alright. I’m no longer going to jump off the bridge. I don’t need to go full on sexy. But I do feel like a tweak is needed.
Now that you’ve made it this far, I’ll admit that I have lied to you a little bit.
I’ve been presenting these doubts and feelings to you as if they’re new. But I also experienced these doubts and feelings the last few years. These aren’t new. And I adjusted slightly during this draft season. Here’s what I did:
- I Splurged on Top Level Players – I even admit to overpaying. A little. Not a lot. I have investments in Mike Trout, Clayton Kershaw, and Paul Goldschmidt. It can be argued that overpaying is a poor idea, but I wasn’t stupid about it. I’m comfortable with it because of the stability, reliability, and proven track records of these players. I think that’s worth something. And we don’t attempt to quantify it. Shouldn’t stats from a player that’s done it repeatedly be worth slightly more than the same stat projection for the player that’s done it once?
Even if you answer no to that question, the other reason I’m comfortable with it is the flexibility it opens up in my roster. If I think I have an edge over my competition by being more in-tune with league news and trends, then I need to open up roster space so I can exploit that edge during the season. Squeezing more of my budget into the top tier players allows me extra room at the bottom of my roster to speculate. I shouldn’t have to sit there this year with my hands tied. Maybe I’m wrong, but I think I can snag a gem out of the early season turmoil that will more that make up for the slight overpay.
- I Pursued Selectively Chosen Under-Publicized “Toys” – One reason going after the “popular” pick up is a bad idea is the competition inherent in securing that player. Demand drives up prices. So you’re probably going to have to overpay. The trick is to search for the unheralded toys. While I don’t have any shares of Socrates Brito, A.J. Reed, or Lucas Giolito, I was able to secure shares of guys like Hunter Strickland, Joe Ross, Vincent Velasquez, and even Byung-Ho Park (who seems to be under the radar in my circles?). I consider these types to be a bit trendy, but not all out.
- I Drafted Injured Players (where I had DL spots to accommodate) – No really an advanced strategy here. But it fits with my desire to open up roster space for early season speculation. This is a double-edged sword (too many injured players can clog up a roster), but the benefit here is even more flexibility to pursue some of the early season wonders. The trick is differentiating Chris Shelton, Emilio Bonifacio (06010!), and Willie Bloomquist from those with legitimate staying power.
I tried to be reasonable here, too. I stayed away from mid-season Tommy John returns and went with the seemingly safer Michael Brantley, Anthony Desclafani, and Josh Hamilton types.
We’ll see how this plays out. But hopefully I was able to identify strengths and weaknesses of mine and have compensated accordingly.
Thanks for the therapy session and listening to me talk to myself.
Tanner writes for Fangraphs as well as his own site, Smart Fantasy Baseball . He's the co-auther of The Process with Jeff Zimmerman, and has written two e-books, Using SGP to Rank and Value Fantasy Baseball Players and How to Rank and Value Players for Points Leagues, and worked with Mike Podhorzer developing a spreadsheet to accompany Projecting X 2.0. Much of his writings focus on instructional "how to" topics, Excel, and strategy. Follow him on Twitter @smartfantasybb.
Another strategy is sticking to your safer bets early (think your mid round talent in a snake) and then with the surplus you’ve saved you can go after a few extra lotto tickets that have potential to pop.
The challenge I’ve found with using projections to drive my ranks in both spring training and during the season, is projection models tend to lag behind the eye test. Sometimes a blind leap of faith is required if you want to secure that next big ticket.
Hey Tim,
Yes. I need to do a better job of considering risk with all picks and coming up with a nice mix. Your point about projections lagging behind is a great one. I try to monitor news to see if an actual change in behavior (new swing, new pitch) corresponds with an increase in performance. But we don’t always have that info.
yea if you want to win, need to take more risks. your strategy is perfect for being above average
you continue to kill it Tanner…I think you’re moving in the right direction.
This is linked to your first article. Your $18 surplus is only surplus if everyone lives up to your projections. But as you noted previously, projections are inherently unpredictable. So your $18 surplus can only go down, while the ‘sexy’ team’s $19 deficit is more likely to go up.
Preseason, buying a bunch of vets at slight discounts results in a projected profit. But the team that wins the league will almost certainly include several ‘breakout’ players. That’s the easiest way to earn significant (league-winning) excess value. Going stars & scrubs opens up the possibility of getting season-changing ‘scrubs’, whereas buying Billy Butler has almost zero chance of big value.
I misspoke. Drafting safe does limit upside, but of your $18 projected surplus doesn’t have to go down.
One other note is that the lottery ticket guys won’t be held all year if they really don’t pan out. By definition, “replacement” level guys are freely available on the wire if the longshots don’t pan out.
You just gave me an interesting thought. I hate being in the situation where I feel handcuffed and can’t drop a player. Guys like Rollins, Lind, and Aoki ARE handcuffs. I’m going to hold them all year and feel like I can’t drop them for the next lottery ticket.
Hmmm. It’s interesting. There’s something to be said for your lottery ticket coming up a loser and knowing you can drop it and pick up a new ticket. It may be a better winning strategy than holding onto my $2 winner all season.
Scenario analysis is your problem.
The playing time scenarios / production – for the veteran guys with your projections, you’re going to know most of the time. Really basic: There is like 75% chance they perform similar to how they did last year, 25% chance they’ll age and do worse. The sexy team is more like 50% chance they’ll do worse (and you have to drop them, and pick up new players… replacement value), and 50% chance they’ll do a lot better.
This is reflective in the pricing of the players. Blake Snell is a good example, he paid $3, if Snell pitches 100 IP, he’s worth $10. If he doesn’t, eventually the guy will probably cut bait with him.
You aren’t going to drop Ian Kennedy because he sucks for a month, because he has a history of being better. Less roster flexibility.
It sounds like you might be miscalculating the replacement level player.
If 20 1B are drafted, the easy approach is to say “ok, replacement level is the 21st 1B.” and then calculate values by contribution above that expected baseline.
But that’s wrong for a number of reasons. First, replacement level isn’t the expected return of the 21st 1B, it’s the expected maximum return of all undrafted 1B. For a single player, the mean of his possible outcomes is a reasonable approximation. But if there are 8 more viable looking 1B, one of them is going to outperform expectations. So realistically, if you have to hit the waiver wire in June, there’s probably a better option then you expected in February.
Then, depending on the frequency of transactions allowed in your league, it’s much easier to swap players around at the replacement level. A 1B on the waiver going to Colorado? Put him in! Somebody’s on a hitting streak? Hot hand is real again, grab him! Again, it depends on league settings, but you can squeeze more value out of a waiver wire position then the stats of any one player would suggest.
Once you make that adjustment, the pool of viable players shrinks and thus their values increase. It also adds value to mid tier high-risk/high-reward players: you get to ignore any outcomes that drop below the new replacement level: if they get sent down to the minors you play the waiver wires, or pick up whoever looks like they might be having an unexpected breakout season and move on. If you’re overly pessimistic on what replacement looks like, the risk on a lottery ticket player looks much more severe then it really is. If you make the adjustment of using a higher baseline for the worst case scenarios on risky players, you might find that some of the sexy picks aren’t as overvalued as it would seem.
That is incorrect. The replacement level absolutely is either the worst player who has to be used or the best player not used depending upon your calculation preference. Upside comes into play when you choosing your players, but it absolutely should not ever be considered when calculating the replacement level. The “maximum return” is irrelevant until it actually happens. Some players expected to be above the replacement level yet ending up below and vice versa doesn’t change the replacement level itself.
It’s subtle, but I do believe the effect is real.
Let’s say there are 4 players at replacement level, all expected to hit 10 HR. Any one of them, the most likely outcome is that at the end of the season they have 10 HR. But as a group, since we’re taking multiple measurements, the most likely outcome is that one of them outperforms and hits, say, 15 HR.
You are correct that this is counteracted by the fact that the most likely outcome is that a few of the players above replacement will fall below it, such that the stats of the replacement player will be 10 HR, even if we don’t expect that at the end of the season that person is who we thought it would be.
However, I do believe that owners closer to replacement level have a “first mover advantage” in that it takes more observations to counteract the priors on a player with higher expectations. If a player is off to a breakout season we slowly accumulate evidence against the performance being a fluke. The same with an under-performing veteran. In general, by the time the owner is willing to cut the under-performing veteran, an owner at replacement level should have taken the best of the predraft sub-replacement players.
This is all very much counteracted by the statistical noise inherent in baseball. But in my experience, it is more common that undrafted breakout players end up replacing very near replacement level players then underperforming lower-mid tier players. So I do think there’s a value “bump” near replacement from the increased probability of getting the best of the projected subreplacement guys.
I understand what you’re getting at, but it has nothing to do with calculating the replacement level. Furthermore, while it’s true that low dollar players are more easily dropped, most leagues will have reserves that provide all teams the flexibility to pursue undrafted players as they emerge.
This is all very dependent on league settings. Large benches and weekly transactions grant a lot of flexibility on rostering prospective players. On the other hand, smaller benches and daily transactions with high IP limits tends to reward keeping starting pitchers on the bench as an expanded roster and penalize using the bench to stash position players for later.
If you’re going to pursue a strict valuation approach, I think a modified replacement level is a reasonable way to deal with the phenomena that if you get the worst player at a position you should get more value then he’s projected to provide. In the end, we win based on the sum of contributions from every position. If there’s value in the flexibility to pursue breakout players or to play strong matchups from a waiver pool, that needs to be factored in somewhere.
The usual approach is to do it ad-hoc, to go with a “stars and scrubs” strategy or to “always take upside in the later rounds”, but if you’re trying to set a value to each player and stick with it, I think modifying replacement production to what you expect a slot to produce over the season playing the waivers smartly accomplishes that pretty neatly.
No, polama, what you’re describing has absolutely nothing to do with calculating the replacement value and should never be incorporated into such a valuation. It is something to keep in mind during an auction depending upon your roster construction, but it does not affect the replacement level.
Why is doing replacement value based on expected contribution from a position selecting from waivers wrong?
And what do you propose instead? “keep it in my mind during an auction” is clearly incompatible with a “go by the values” approach. If you don’t do the calculation in replacement value, where do you do that calculation?
Polama, I don’t agree with your method of determining replacement level, but it may not even matter. Regardless of what replacement level is, the players at that position are still going to come out ranked in the same order. So it wouldn’t yield any different results.
But your comments about the danger of owning veteran players that are “slightly above replacement” is well taken. That’s a common experience for me. I can’t drop them and I miss out on opportunities to cycle through potentially more valuable adds during the season.
I had the same issue last year with too many boring players and it turned out poorly compared to my expectations. Other teams who were aggressive and bought into the sexy prospects and one-year wonders (guys like Bryant, Sano, Syndergaard, Turner, Arrieta, Keuchel) reaped the benefits.
This year I again built my entire draft board using projection-based values, and then one by one I go through the list looking at age/experience/injury situation and add or subtract upside-based value entirely on feel. I try to throw in a significant negative to all players dealing with injuries in spring training, having had bad results with such players and by the first week chances are my DL slot will be full anyway with a formerly healthy player. The main place I tried to focus in rearranging my draft strategy was in the mid/late rounds, where you can gain the most value by taking shots on upside and where misses are more replaceable. Plus, as we all find, having players who miss and need to be replaced is not the worst thing, as like clockwork every year new faces will emerge, unknowns will break out, and closers will be shuffled.
The seasons hardly started but I still feel good about my draft now. When you add value based on feel there’s risk of poor optimization, but I loved minimizing that dreaded regret factor, and it makes the game that much more fun by buying the guys you believe in. I trust my judgement to lead me right more than wrong, might as well plant my flags to it.
Why are you playing fantasy baseball? If it’s primarily about winning then you’re doing the right things by trusting your prices and looking for values wherever they present themselves. However, if you derive more pleasure from “owning” interesting players and following them during the course of the season then your approach could benefit from putting more emphasis on acquiring those names even if it hurts your chances of winning. Personally, I don’t get all that much satisfaction out of winning leagues. My league trophies sit in a corner gathering dust until I have to take them back to the auction next year. What excites me is having a player I liked more than most validate my opinion by having a breakout season, or even sticking with a player who’s struggling and fanatically digging through box scores and FanGraphs data to find reasons to believe they will turn it around. I still try to win, so it’s not as if I’m picking 23 names before the draft and stubbornly bidding them up to whatever, but knowing that winning is not the most important thing allows me the freedom to deviate from my dollar values.
One thing to note is that if you do spend on high dollar players then it makes sense to complement them with “boring” low dollar picks to mitigate some of your risk, whereas a team centered around second and third tier players is a little more free to gamble on upside when filling out the remainder of the roster. Similarly, it’s a better idea to gamble on upside in mixed leagues where you have many more options for replacement than in single leagues where the free agent pool is incredibly shallow.
I find you always need a gamble or 2 to hit if you want to win a league. If you don’t hit, someone else probably will. Try to build a nice base and then round out with upside.
In my experience the opposite is true, the teams who “gamble” in the auction are far more likely to go bust. We tend to remember the rare successes while remembering the much more common failures. I would argue that what’s important for winning is to supplement a solid, value-based draft with upside in the reserves and being on the ball with free agents during the season.
err … forgetting the failures.
Agree with Nate here. The strict value owners never win in the leagues I’m in. They usually finish in the top half, but never win.
I love this conversation that Tanner has started. I hear a lot of experts (and commenters) tout the strict value approach, but I never thought it worked and it seems that a lot of others agree. Kudos to Tanner for challenging the “accepted” practices.
I find it extremely hard to believe that you’ve never seen someone with a strong value-based approach win a league. Betting on breakouts is a good way to lose because you’re gambling on long odds, so even if you hit on one of your gambles you’re likely to miss on several others. Where some “strict value owners” fall short is that they don’t adjust their values during a draft to account for their competition, and therefore you might end up very strong in certain categories that the league undervalued but weak in others that the league overvalued.
As with all things, I think the key is balance. Some risk, some safety, some upside, some floor. All nicely mixed together. I think I’ve been too far in the safety and floor side of things.
And it’s fairly hyperbolic to say strict value users never win. Larry Schechter wins all the time. Against the best of the best.
Those of us that aren’t Larry may need to be more flexible and mix in some risk and upside.
One way to handle this is to pursue different drafting strategies for your starters and your bench.
If you want to go conservative (read: value) for your starters that’s fine. It’s probably safe, and will insulate you from crashing and burning. It’s boring, but will give you a decent floor to work with.
But shoot for the moon with your backups. In fact, my philosophy on bench players is encapsulated by one word: “upside”. I have zero interest in carrying a scrub 4th or 5th OF who is not significantly better than the guys left on the waiver wire. I can go get them any time I want to. My bench is always made up of breakout picks. If they bust, so what? They’re just my bench and I can reload as needed. But if they break out then I’ve got a star _and_ I can trade my boring safe starter for something else.
RE: your sexy v boring list in the paragraph after “Is There a Weakness in My Game?”
Other than JOsh Donaldson, the sexy list is young guys on the right side of the aging curve. Your boring list is all on the wrong side of the aging curve. I decided awhile ago that I wasn’t going to draft older players anywhere near the top of the draft unless they were so obvious I simply had to & the obvious ones never fall to the point where I would take them so I basically never draft old guys at the top of my draft. Note that Donaldson is now out with the same kind of old-guy calf strain injury that hit Miguel Cabrera last year – in both cases the guys pulled a calf just simply running.
I ended up drafting only 3 guys over 29 – Beltre (defying father time and still a good 3b option in a league that has CI), Lind and Victor MArtinez who were both very end of the draft value plays that will probably get dumped before the season is over for an AJ Reed or someone similar.
I have gone through similar thoughts the last few years. The goal of the game is to give yourself the best chance to win as opposed to getting the most value at the draft. If you draft value based boring teams, you end up like the Saint Louis Blues (25 years in a row in the playoffs but no Stanley Cups).
The goal at the draft should be to get a combination of getting value, upside and roster flexibility.
The 1st change I made was avoid taking “boring” players in the late rounds of drafts unless they were very undervalued. For example in a 12 team league( 5 OF per team ) say I have Matt Holliday ranked as my 50th OF (160th overall). I would have pick 180 and think to myself I got value. If he performs as expected it is a slight value. (Since you need over-performance to win a league, does he do enough?). The bigger issue is when he under performs. How long do you wait to drop him? He clogs up a roster spot and you miss out on the early season free agents.
If you draft a Marcell Ozuna instead you essentially giving yourself two ways to win. If he breaks out then it is a big win and if he doesn’t you don’t feel as attached to him and you can search the waiver wire. There will be a number of times when the “Matt Holliday” outperforms this situation, but it is a smart risk to take if you want to win a elague.
This is said often in DFS, but not enough in season-long. But you’re right. I need to consider this more often. And I like the Ozuna scenario you describe.
Really interesting read, thanks for sharing. A couple of thoughts:
There are two flavors of $1-3 guys. One is the Jed Lowrie/Chris Iannetta/Ian Kennedy/Erasmo Ramirez type – pretty well-known commodities, often veterans, who are fairly projectable (barring injury). If everything goes as expected that $1 guy will earn you $2.56 or $3.23 or something, but there is very little chance that these guys “break out” to a totally unexpected level and earn you $10 or $12.
Then there’s the Vincent Velazquez/Tony Zych/Socrates Brito/Trevor Story type – guys who may earn you little or nothing (in which case you can just cut bait and pick up someone else) but guys who could also “break out” and earn you $8 or $12.
Group A is the safer group; they have a higher floor and will generally earn you *something* positive. But they won’t win you the league. There will be plenty of busts ($0 or worse earners) in group B, but there will also be some guys who earn $10 or more. Stumble onto a few of those (in addition to your studs) and they might win you the league.
Or, more succinctly, no one won a league by getting Howie Kendrick for $1. But you just might by throwing $1 Jeremy Jeffress, Luke Gregerson, or Socrates Brito. And if Jeffress ended up relegated to 8th inning duties? No big loss, just cut bait. But the group A players have juuuust enough value that you have a hard time cutting them loose, especially to pick up a guy who may be in a set-up role (or the minors) in two weeks.
I would argue that your APPROACH is right on, and the flaw (if there is one) in your overall strategy is in the projections you use when executing your approach. I find that people who discuss value projections tend to assign FAR too much certainty to those values, and miss out on the fact that baseball is neither a spreadsheet sport nor a skills-based sport, but a hybrid of both.
I’m an advocate of what I call “profile-based evaluations”. In other words, I start every given fantasy baseball season by getting a sense of the projections of the consensus, and then apply my own baseball knowledge on a case-by-case basis to determine which consensus market values are over- or under-valued.
To use one example, look at C this season. I wasn’t too keen on the few top investments at the position, and then the remaining options are a mass of similarity and risks. So my goal was to invest in a lower-priced guy who I had confidence had much higher odds to exceed the projections that his price was based on that others. One of those was Blake Swihart, whose price was basically built around taking his 2015 line and regressing his very-high BABIP. What most people didn’t factor in was 1) his 2015 line took a HUGE hit from Swihart’s miserable first 15 games, 2) his BABIP could be expected to be higher than that of most catchers, and 3) there was reason to expect more rapid growth from Swihart than many others.
1 was based on the reality that Swihart was rushed to the majors on an emergency basis. At the time of his callup, I EXPECTED misery from him, and he did just that. What I didn’t expect was that it would only take him 15 games to adjust to the MLB game, and almost overnight he became a different hitter for the rest of the season. In other words, the early struggles were both absurdly bad AND COULD BE ESSENTIALLY TOSSED OUT as the product of a premature callup that had since been completely overcome.
2 was based on the realization that Swihart consistently posts higher BABIP figures, and on the fact that that result is intuitive given that Swihart has high athleticism for the position and has a good contact/line-drive approach.
3 was based on knowing Swihart’s history as a player who hasn’t had a lot of professional exposure, is still very young, and has shown fairly rapid growth at all levels despite the handicap of his entering the professional game with VERY little catching experience (developing as a hitter is hard for almost all catching prospects, and most of those have played the position long enough to not need to have devoted as much attention early on to defense as Swihart).
And so, my point: look at the amount of QUALITATIVE analysis that went in to my determination that Swihart offers a significantly return than his market value. And I try to determine such things (positive OR negative) for as many players as possible. I say that you need to build in some degree of understanding of these factors into your valuation approach, if only to get a sense of player upsides, downsides, and the odds of each player ending up at different points of their possibility spectrum.
I am team Sexy and have some of the opposite problem. You see I make my own projections by looking at the average of 3 different projection systems, career averages, age, personal bias and then make my own projections based off of that and put it into excel and base values around my own projections. Problem is sometimes it gives you crazy results like Roughned Odor ranked as the 2nd best 2nd basemen. When looking at these I do adjust for their actual values by just studying them and knowing not to overpay to much during the draft (auction) but I usually end up with a lot of upside and some risk on my teams.
Can’t stop laughing at this.
I am in a simple POPDFL (Pick offense pick defense fantasy league) pool.
At your turn at the draft for hitting you select a team only for hitting. At your turn for pitching you select a team only for pitching.
The game number for each of your two teams have to match up throughout the season. For example, I was able to get the Cubs offense and the New York Mets pitching. I write down what the Cubs scored in their game number one. Then write across that I write the amount of runs the Mets gave up in their game number one.
If the Cubs score more than the Mets have given up it’s a win for me. If the Mets give up more than what the Cubs have score in the same game number it’s a loss. In our pool, a tie is a win.
Our people enjoy the team concept more than the individual player concept.
However I must say this article was greatly enjoyed.
Confirmation bias is rampant in these comments. I’d say you’re just overthinking it. You are drafting the best team with the best value based on your projections, that is giving you the best chance to win. You seem too worried about the names tied to those projections.
Thanks, Descender. It’s worth considering that the process is not flawed. Bad results don’t necessarily mean a bad process.
I’m not overhauling things. But I am tweaking. Hopefully it’s an incremental improvement and not a step in the wrong direction.
SCREW Value in season LONG ROTO – if u r trying to win this year…yeah, value is great but u GOTTA HAVE STATS
P.S. – No matter what “BAD RESULTS” result in NOT-WINNING
I used to do your kind of draft. Another guy in the league used to call me “The AARP Team,” because I was always loaded with solid veterans. A couple years ago I made a conscious decision to go younger. It’s more fun now. Maybe you should try a variation and use your projected values but commit to not taking anyone over 30 years old. You’ll end up with some big, young stars at maybe not as much of a value savings, but more speculative young guys. Just do it in one league and see if it isn’t more fun. Part of the fun, for me at least, during the season is spotting gems off the waiver wire. Going younger lets you do so when some of your young guys don’t pan out.
There are three problems with using projections for roto: First, the heavy regression used makes it near impossible to predict a breakout player. Second, projection systems seem to weight stats from an injury plagued year as normal, instead of giving the player a pass for a performance which was bad for obvious reasons. Third, many “advanced” stats may be good for things like Hall of Fame or MVP or Cy Young arguments where you try to find the true value of a player, but in Roto you don’t care about the true value of a player, you care about the real life value. Unless a player changes teams (and you have to guess about his new situation) you should never use Park Adjusted stats. Unless a pitcher changes teams you shouldn’t put as much weight on Fielding Independent stats. A pitcher comes with his fielders in roto. Upgrade pitchers with good fielding behind them and downgrade those without. Upgrade batters who are in good lineups because they will have more runs and RBIs. Downgrade batters in weak lineups. Upgrade guys who move near the top of the order, downgrade guys who move to the bottom. Grab a couple guys who had outrageously good springs. In short, look at all the Bold Predictions on RotoGraphs. Some of them are really a stretch, but make some slightly less bold predictions of your own based on things the projection systems don’t capture, use them in your draft, and you’ll have a lot more fun in your season.
I think this philosophy plays out better in a snake (non auction) draft. The best predictions are only in the 70’s for accuracy and a $ value from those really should not be so rigid, especially when your opponents have different values and strategies. If you play around with any generic $ value software the values change (sometimes dramatically) just by changing the H/P spending breakdown and you will see it more so with the high end and “name” players I bet.
You can get some really good value picks late in the draft and with certain de-valued players or positions. I’d keep with your strategy to prevent significant reaches or over-spending, but sometimes you need to maximize your other “value” picks with some solid high end or “sexy” names that have better over-achieve possibilities.
Jimmy Johnson gave Bill Belichick advise about drafting which they both used to great success which can be applied to this. Make a realistic list of players you really want and go hard (not at all cost) for them. Your draft board and value system will fill in the rest.
I play in a points league, and this is my second year playing. My experiences also come from two years of playing fantasy football as well. In fantasy football, I went from bad luck/out of the playoffs, to league champion, despite losing six of 9 starters for the championship. Baseball, I finished fourth, but it’s a keeper league, and I had a strong set of keepers. (Six keepers, 25 players + 5 bench + 2DL, with 2OF , and 3 UTIL)
When it comes to a points league, Clayton Kershaw is pretty much the great equalizer you can have on your team. I noticed during this week, where I was playing an inferior opponent who kept the game a lot closer than it should have been, that I also had a team that was much too “safe”. So, for going into next week, I dropped some weaker bench guys for guys with more upside. Some guys I have plugged into my lineup already, some are moves that are meant to keep potential impact talent out of the hands of my opponents.
I also pulled off the very rare, week 1 “challenge trade” (trading younger, potentially impact guys who are currently struggling for each other). Randall Grickuk for Wil Myers. My team is VERY strong at 2B, C, SS, and 3B. 1B was my one worrisome spot. Basically all the good 1B were kept. That’s how I decided to draft my team the way I did. I know it’s different without actual dollar values, but the value I got in the draft was MASSIVE. Three of the weakest positions, I got more than one above average player.
I think for people like you, something like the “challenge trade” is the best way to get past the fear of dropping a guy you like. When me and the other owner made our trade, we even said that one of the best things about that trade, besides the fact that I could use some help at 1B, and he needed depth in the OF, is that it’s a lot easier to drop guys that you traded for, rather than guys you drafted. It’s inevitable that we’re going to value the guys we drafted higher than guys we traded for or picked up off waivers/free agency. If Myers flops, he’s easy to drop.
This is a strategy that helped me win in fantasy football last year, and I also employed it in baseball last year, and already this year. Since I have 5 bench spots, plus potentially two DL spots, I try to make the best use of every single roster spot. If a team drops a good player who got hurt because their DL is full, I’ll drop a bench guy, put him on my DL, and pick up another guy for my bench. But in the end, in a points league, and I would definitely think in roto as well, if you want to win, you need to keep potential impact talent away from your opponents. This is especially true if you have weekly lineups like we do in my points league. It’s always better to drop a guy you’re not using, than to let your opponent pick up talent for free. If they want to pick up someone I’m dropping, go ahead, you need to give up your waiver priority to do so (waiver priority never resets in our baseball league; it reset from worst to best record in football). And, if they’re not good enough to be every week starters for me, then you’re easily droppable.
I guess I’m a bit luckier in that I don’t get quite as attached to draft picks as other people do. Both in baseball and football, I am RUTHLESS in adding and dropping guys, because you have to keep talent away from your opponents. It’s rare that I’ve ever felt burned by dropping someone for a better bench player, or as a speculative add. I’ve also noticed in trade talks, that some people complain that they’re not getting enough back, because they drafted player X in round 3, and they’re getting player Y I drafted in round 10 back. After the draft is over, where you got drafted is irrelevant to me. A trade is one of talent for talent, not draft slot for draft slot.
I do admit, though, that teams and/or players that I do want to watch, there probably is a slight bias towards those players when I draft, trade, or pick up off waivers or free agency. This is especially true when it comes to football, but does apply in baseball as well. I think that it’s important that you do have players from teams that you want to watch. I have MLB.TV, and live in Canada, where there are no blackouts at all, so I can watch every single game. Value should come mainly through production, but you can’t completely ignore the eye test and entertainment value. To me, fantasy is supposed to ENHANCE the enjoyment of the sport. Talent will always come before the team someone plays for (not a Yankees fan, but I have bunch of their players, for example), but if you’re doing things in an entirely mathematical sense, then you’re losing out on the entertainment value that can be gained.
And I don’t know how much trading happens in your league, but ours, there’s been a few blockbuster trades, and a lot of smaller trades as well. Football… not so much. But for roto, even moreso than a points league, I think you’re missing out on the hidden value of loading up on weaker positions. If someone goes down with a major injury, you now have yourself a trade partner. At an especially weak position, you can generally extract a higher value than you’re sending back. Surplus value doesn’t only have to come from the draft. You can acquire it through trades, and by wisely dropping bench guys to keep talent away from your opponents.
I will finish with this though… I do have my horses for my starting pitching, but there’s two-to-three guys per week who I’ll stream. I don’t see that changing at all. But my offense… Even after the Myers trade, I got some additional 1B depth, plus great depth and power in my OF as well, all for free from free agency. I made some bets around the same time last year, and they worked out great for me. But it’s gonna be harder for me to speculatively pick up offensive guys this season, I think. But that’s when you gotta try the trade route first. And if it’s someone who could be a keeper, versus a above-replacement level veteran, I’ll bet on the potential young keeper.
But in the end, a lot of it has to do with my roster construction. When you have a guy like Kershaw who can often single-handedly cover up a roster hole for a week, you can bet on upside more than floor a lot easier. And having the best players at the weakest positions already gives you a better floor to begin with. I think the biggest difference we all have is what we have as an acceptable floor. C, 2B, and SS all have very low floors after the elite talents. 1B and OF have pretty high floors, OF especially. In my case, if 1B is my weakest spot, maybe it’s replacement level, but the rest of my team is so far above everyone else, that doesn’t even matter. To me, at the weakest (or, in football, the most volatile, RB) positions, you want the highest floors, the elite talents. I think we generalize our entire roster as a single “floor” and “ceiling”, which I think is incorrect. I’ve read articles about the “myths” of positional scarcity, and while I wouldn’t say they’re wrong, the 1B and OF floor are a lot higher than those of C, 2B, and SS. Make a good trade, and you probably get back any excess value you paid to get that player, and then some.
I do like that you linked to that ESPN article about not derailing your draft. I thought I did in football this year. It was a truly headscratching draft, and in a bunch of rounds, there were runs on players at certain positions. I joined in on those runs. I HATED my team. I had an ok start, but ruthless add/drops gave me the best, deepest team in the league, and I won. There were runs in our baseball draft this year, but even before I read that ESPN article, I never let those runs derail my draft plan. I saw those runs as an opportunity to load up at weak positions. Now, if anyone has major injuries, their trotting out trash, or they have to trade with me. There’s no better place to be than there.