Steamer and I: Corey Dickerson & A.J. Pollock
You know that I develop my own player projections, of which I have lovingly dubbed Pod Projections. With over 500 players projected, I thought it would be interesting to compare them with the Steamer forecasts. So Steamer and I is a new mini-series in which I discuss a pair (maybe more, maybe less) of players who I have projected for dramatically different results than Steamer. To do this, I simply compared my wOBA projections with Steamers’, calculated the difference and then sorted.
To begin, let’s discuss two youngish outfielders who enjoyed breakout performances in 2014. I am projecting both of these outfielders much more optimistically than Steamer. What has led to my bullishness? Let’s find out.
To make comparisons simpler, I will extrapolate the Steamer projections to the same number of plate appearances that I have projected.
| System | 2B | 3B | HR | BB | K | AVG | OBP | SLG | wOBA | ISO | BABIP |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Pod | 31 | 6 | 26 | 41 | 111 | 0.298 | 0.350 | 0.537 | 0.382 | 0.239 | 0.335 |
| Steamer | 29 | 6 | 22 | 37 | 107 | 0.285 | 0.334 | 0.499 | 0.360 | 0.214 | 0.321 |
Let’s begin with what Steamer and I generally agree on. We both project an identical number of triples. We’re also fairly close on both walk and strikeout rates, though Steamer is forecasting slightly lower rates for both metrics. The doubles projection is also relatively close, but still counts as the first data point contributing to my more optimistic ISO mark.
We we find the largest gaps are in home runs and BABIP. We’ll talk homers first. I’m projecting a 17.5% HR/FB rate, versus a 19.5% mark in 2014. Unfortunately, Steamer doesn’t project HR/FB rate directly and there’s no way to calculate it without a batted ball distribution projection. But here are the facts — Dickerson’s 299 foot average batted ball distance ranked 15th among hitters on my spreadsheet and his 21.7% xHR/FB rate ranked ninth. And he consistently posted ISO marks well above .200 during his minor league career. Although his Triple-A performance came in a seriously hitter friendly environment, he still calls such a place home, it’s just in a different location. I did choose to project some HR/FB rate regression, but not to the degree the Steamer projections are implying.
Now how about his BABIP? Dickerson posted a monstrous .356 mark in 2014, which any projection system is going to regress heavily, especially involving a player with limited Major League experience. But his xBABIP was even more ridiculous at .374! That’s what happens when you combine immense power, above average speed and a 26.7% line drive rate. He also typically posted inflated BABIP marks in the minors. And given that Coors Field boosts line drives by 18% with its 109 park factor (easily the highest in baseball), it suggests this batted ball profile shouldn’t automatically be expected to significantly worsen.
Let’s also not forget that Dickerson doesn’t turn 26 until this season, so if you just look at his age, he should still be in growth mode.
| System | 2B | 3B | HR | BB | K | AVG | OBP | SLG | wOBA | ISO | BABIP |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Pod | 33 | 7 | 13 | 39 | 87 | 0.283 | 0.336 | 0.450 | 0.343 | 0.167 | 0.320 |
| Steamer | 26 | 6 | 9 | 34 | 88 | 0.260 | 0.310 | 0.393 | 0.311 | 0.133 | 0.296 |
Out of every hitter that I projected, Pollock had the second largest positive differential between my wOBA forecast and Steamer’s (Scott Van Slyke ranked first). We agree on his strikeout rate and are close on the triples, but that’s it. I’m projecting a stronger walk rate, and a better mark than 2014 for several reasons. Pollock doesn’t swing at too many pitches outside the zone (below average O-Swing%) and also doesn’t swing at all pitches as frequently as the league (below average Swing%). At the very least, it was enough to assume improvement back toward his career walk rate, also factoring in that walk rate should still be expected to rise at Pollock’s age.
Once again, the big difference stems from the power and the BABIP. First, the doubles really jump out. Pollock was a doubles machine last year, launching a double every 13.95 at bats. For his career, he’s at 15.47, which is right about where I projected him (15.5) this season. I’m not sure what Steamer is using to forecast a much weaker AB/2B mark of 19.0 as it doesn’t fit in with Pollock’s history at all.
Moving past doubles and we land on home runs. Pollock’s HR/FB rate jumped from just 6.6% in 2013 to 9.5% in 2014. Surely that is what Steamer is looking at when projecting big regression. I, on the other hand, am projecting a 9.0% mark, as his batted ball distances these last two seasons have sat just above the league average, as has his SDD marks. His 2013 xHR/FB rate came in at 12.0% and 2014 mark 10.3%. Since he also plays in a home park that sports a right-handed home run factor of 105, I’m quite comfortable with my home run forecast.
Last is the BABIP, which I am seemingly more optimistic on for 99% of players compared to Steamer. Pollock posted a .344 mark in 2014 and .314 in 2013. However, his xBABIP hints at a bit of good fortune, as those marks sat at .308 and .303, respectively. And yet, I’m projecting a .320 mark! First, Chase Field boosts both doubles and triples and park factors are not accounted for in my xBABIP formula. Second, Pollock has posted high BABIP marks in the minors, which suggests perhaps he owns a skill not incorporated into xBABIP yet. And last, his .308 xBABIP in 2014 was partly driven by a poor 14.2% line drive rate, which I am projecting to rebound to 18% this year. An 18% line drive rate would have pushed his xBABIP up to .322. Even if I may be just a tad bit high on the BABIP projection, Steamer seems far too low with no rhyme or reason behind its projection.
So which side do you lean toward, mine or Steamer, for these two multi-skilled outfielders?
Mike Podhorzer is the founder of ProjectingX IQ, an advanced fantasy baseball analytics platform that transforms projection data and in-season performance signals into actionable intelligence. He is the 2015 Fantasy Sports Writers Association Baseball Writer of the Year and three-time Tout Wars champion. He is the author of the eBook Projecting X 2.0: How to Forecast Baseball Player Performance, which teaches you how to project players yourself. Follow Mike on X@MikePodhorzer and contact him via email.
Good stuff Mike.
Sorry but don’t know you guys well enough. Met Eno in Chicago. Who does Streamer, Sullivan? Have you two discussed his projections? Streamer seems to be lower than everyone on most projections. So you think he needs to tweak his calculations? You also use more data than he does. Do you think he needs to use more variables? I know it ain’t easy and I’m not trying to knock them but wonder if they need a tweak.
Steamer was created by Jarred Cross and some others. Check their site at Steamerprojections.com for background.
I side with you on Dickerson and Pollock. I was lower than you on Tomas.
I think you nailed Dickerson. I think Pollock is going to split the difference between your projections, mostly because you’re a little high on the BABIP, while Steamer is a little low.
I’m curious about your stolen base projections for both guys, but particularly Pollock, since I’m targeting him this year.
Also, you mentioned that these numbers are based on the number of PA you have projected, but you didn’t note how many you project.
As always, good work!
Yeah I didn’t include the PA projection because I was focusing on the performance ratios and what has led to the optimism. I think I’m at 550 for both. I’ll include next time. The SBs will have to check when back home from the office.
20 steals for Pollock, 9 for Dickerson.
I’m a bit late on the reply, but thanks!
I will chose your Dickerson projection to earn some brownie points.
Probably the biggest factor for his projection is how many PA he receives against lefties. He is not awful against them but it will drag down the BA.
I think he beats Steamer HR projections either way.
Each of the Rockies’ starting outfielders are LH bats. The only player of those three getting platooned will be Blackmon. Gonzalez and Dickerson are everyday bats, no matter who is pitching.
I’m am thinking about keeping Dickerson for $12 in my 12 team mixed league ($260 budget). I know it’s a bargain but I do worry about his righty – lefty and home – road splits. Last year he was basicially a 4th OF or worse against lefties and on the road. Last year only 20% of his plate appearances were against lefties. If he no longer is platooned, he should have a higher percentage of his PAs against lefties (30-35% of MLB pitchers are lefties).
So, in a H2H league, I fear that he will provide no more than a 4th OF value in over 1/2 of his games (closer to 3/5) even though he basically is the best player in baseball at home against righties. My other keeper options for my final sport are: Jake Arrieta $7, Mookie Betts $7, Todd Fraizer $10. Thoughts?
Assuming that you’ve got roster space, you can always platoon him, even if the Rockies don’t.
Who are you keeping that you cannot keep both Frazier and Dickerson at those prices?
13 team 6×6 (OPS and QS) Roto – 2 C, 1B, 2B, 3B, SS, IF, 4 OF, 3 U, 4 SP, 3 RP, 2 P, 6 Bench, 3 DL
I can keep up to 4 of these players for just this year while forfeiting the provided round’s pick:
Jose Abreu – 6
Jose Altuve – 7
Corey Kluber – 16
George Springer – 19
Corey Dickerson – 26
Dickerson at 26 is overpriced, no value there. Take Springer at 19, he’s got more power/speed upside and younger. Altuve and Abreu are obvious.
How is Dickerson overpriced?
Just realized you probably thought those were $ amounts
I’d ditch Springer–I expect he’ll provide similar value to Dickerson, but at a higher price point for you.
This is a good problem to have. I would take Dickerson in the 26th over Springer in the 19th, but that’s really tough and they’re both crazy bargains.
No way you can cut either Dickerson or Springer.
It has to be either Kluber or Altuve that you ditch.
Fans are naturally optimistic (not necessarily a bad thing) so, given two projections, we will almost always side with the more optimistic projection for each player. We consistently find players whose projections we find two pessimistic and rarely find players whose projections seem to optimistic — even though the projections on the whole are unbiased.
I tend to consider fan projections closer to a ceiling, with steamer representing a floor. That reasonable?
Steamer should be a mean projection and the Fans are (generally) are a bit higher than the mean. I’d argue that ceiling and floor basically don’t exist.
Please make that briefly make that argument.
Cheers!
I’ll try, even though you didn’t ask me. Every player’s floor is zero, because they could get hurt. I don’t know the exactly probabilities, but maybe the chance is like 5-10% for position players and 15-20% for pitchers. Realistically there is something of a ceiling in the neighborhood of 7 WAR. Going above that is really hard. But just look at Chase Headley or Matt Carpenter or Chris Davis or (maybe) Josh Harrison and Michael Brantley and Jose Altuve for guys whose true talent is in the probably in neighborhood of 3-4 WAR or less, but hit a “ceiling” of 6-7 WAR in the past two or three years. Any 2-3 WAR player could post a 5+ win season – probability suggest it will happen a time or two a year – but picking who it will be is really difficult.
Yikes, *too*, not *two and *to*. I need that morning coffee.
Anyway, I should have added, nice job walking through the differences and what might be causing them.
I’m curious what you gave Dickerson for runs and rbi. His rate stats last year were ridiculous, and I know you can’t extrapolate too much, but if those numbers are in the 90’s and you throw in 8-10 steals, doesn’t he have 2nd round value with a 4th-5th round ADP?
Not quite that high, but yeah, my projections say he’s a bit undervalued.
I’m disappointed, as upon reading the title I thought there would be some comparative analysis with Withnail and I.
Mike, are you worried at all about a possibly crowded outfield in Arizona? Would you hesitate on Pollock or Trumbo/are you dropping their PAs down at all? Thanks.
My current PA projection for Pollock assumes Tomas at 3rd, and even then I’m only projecting 550 for him. It means I don’t expect him to start every game given their depth, and that would come down further if Tomas can’t cut it at 3rd.
Right. I am really referring to Tomas not cutting it – which I hear more and more might be the case. And I was just wondering if you also thought if Tomas got ABs in the OF if Trumbo would lose any.
There is certainly risk Trumbo could lose a bit of PT, since he’s poor defensively in the OF.
Wow…I wish I owned Dickerson so I could sell at these prices. I say Steamer is closest, though an optimistic BA (will be closer to .265 at years end IMO).
I hope Pod is right, though I fear the regression. Either way, for $5 keeper salary in a 250 cap league (10 team NL) I’m holding on to him to find out.
Keeper conundrum- Roto deep lineups (including 4 OF, 2 UTIL) in a 12 team H2H points league (no minus for K) where we have an 8 start SP limit. Obviously in points leagues SP are the high scorers. Do I keep a $5 Dickerson or a $5 Carrasco? How do I choose between my man crushes? For context, each teams keep up to 5 players and I already have Abreu, Donaldson, Kluber, and Harvey as cheap keepers? Thanks!!
I can keep 2 out of 3, Dickerson, Betts, or Soler and lose 15th and 16th round picks. Was originally leaning Betts and Dickerson but now leaning Betts and Soler.
Thoughts on Soler v Dickerson?
Mike, I’m wondering if you compared the projected wOBA for your projections and for Steamer for the entire hitter’s pool to see if there’s a big difference there. Is it possible your projections for all hitters are higher than Steamer’s? Did you adjust them to the same baseline?
I did and Steamer is more pessimistic. Taking the average (without weighing for plate appearances), I was at .323, Steamer .318. Not a big difference though, but meaningful. I didn’t adjust anything because I’m just looking at the largest gaps on both sides.
Totally stoked to see Dickerson and Pollock together in the same write-up. I have both in an NL-Only ($280 Cap) at $4 and $5 respectively. I can’t tell you how high I am on both of them. Would love to get a combined 40 HR and 35-40 SBs out of them this year.
I hope to see Dickerson get some ABs in front of TULO and Pollock in the two-hole in front of GOLDY (also on my team at $38). Giving these guys choice pitches would do wonders for their stats.
Dickerson is a fascinating hitter. Probably has a wider than average reasonable variance of outcomes for 2015, depending on his ability to improve versus LHP, Weiss’ confidence in him (Weiss seemed to favor Blackmon until August – Dickerson outhit Blackmon from early March on), and his ability to make adjustments as the league adjusts to him. He seems to have far above average ball-striking ability – he’d get down 0-2, 1-2, and still find a way to battle back and put together a quality at bat and create hard contact. I wouldn’t be shocked if he hit .250/.310/.425, or .300/.350/.525 with 30 home runs.
I have faith in Dickerson hitting lefties. He’s hit them at every level of the minors.
2010: .299/.356/.582
2011: .245/.312/.588
2012: .342/.390/.596
2013: .293/.340/.525
Dickerson’s home/away splits are pretty massive, for what it’s worth (even for a player in CO). What that means in terms of how likely it is he’ll regress, I have no idea.
Rockies players tend to have their H/R splits even out the longer they’re in Colorado as they get used to adjusting to the Coors Hangover.
Love your articles because you always seem to give a very good basis for your arguments and I learn a lot through that.
Did I read somewhere that LD% isn’t something that should be used in predicting other stats?
Also, do the pod projections include a projection of errors? If not, can they? Believe it or not our league uses errors as a fantasy stat.
Thanks and yeah, I always, always have a statistical basis for my opinions. LD% is certainly a batter skill, not so much for pitchers.
Nope, I have never projected errors before and unfortunately don’t ever plan to! I’ve never read any research on error aging curves so wouldn’t even know how to project it aside from just a simple 3-year average or something.
If I’m a fantasy owner, I worry about Dickerson’s ceiling, because he’s never going to get middle-of-the-lineup ABs when the Opening Day starters are playing. Despite all his ability, he may only the 4th or 5th best hitter on a really stacked team.
That said, I expect his BABIP to regress to the mean, but I also his road numbers to regress to regress to the mean as he enters his 2nd season — offsetting any fall-off. Dickerson has raked at every level. He’s just a natural hitter of the baseball.