Minors to the Majors: Hit Tool Grade Usefulness
Earlier in the offseason, I examined out how reported Hit tool grades compared to actual MLB batting averages. I called the process a “mess” but figured it had some value. When I implemented the formula on MLB.com’s 2017 grades, commenters had the following to say about the projected batting average values:
“… not enough differentiation there in my opinion”
“… adjust your outputs to create more difference..”
“… hoping the table would be more conclusive…”
“…way too tightly grouped to the mean…”
“…it’s better to have no projection than to project everyone to be average…”
“… regressing too much to the mean…”
“… hit tool grades should be ignored…”
“…hit tool is undervalued in prospect analysis…”
I have no issue with the hit values being regressed to the mean. What I do have a problem with is if the hit tool is not measuring the correct factors. I needed to find out if reported hit grades provide any value. The following is a detailed look at how the hit tool is graded and how it fails to predict one simple factor, a hitter’s ability to get hits.
Industry standards
The hit tool is probably the hardest tool to evaluate as Kiley McDaniel explained in a six-part series a couple years back (Part 1 with links other five articles). To start with, here are some Hit tool definitions from various publications.
• 2017 Baseball America Handbook: Hitting ability is as much a skill as it as a tool, but the physical elements – hand-eye coordination, swing mechanics, bat speed – are key factors in the hit grade.
• MLB.com: No definition.
• 2080’s Dave DeFreitas: Bat speed, bat path, exit velo, command of the [strikezone], ease of actions, balance during [the] swing.
• BaseballHQ’s Minor League Baseball Analyst: Measures the player’s ability to hit for batting average and judge the strike zone.
• Kiley McDaniel’s three aspects to the hitting grade.
- Tools: “This category includes bat speed, raw strength and the basic structure of the swing.”
- Bat Control: “…ability to change your swing to match the pitch that’s being thrown.”
- Plate Discipline: “… be able to recognize different types of pitches, get yourself into good counts, lay off the slider in the dirt, etc.”
While most of the prospect grades I have collected have been from MLB.com and Baseball America but they provide the fewest details … if any on grading the Hit tool.
Top Hit grades
With the Hit grade “defined”, the following table contains hitters with a 70 or higher Hit grade (players are only listed once if on multiple rankings).
| Name | Year | Publication | Hit Grade |
|---|---|---|---|
| Oscar Taveras | 2014 | BA | 75 |
| Andrew Benintendi | 2017 | BA | 70 |
| Austin Meadows | 2016 | BA | 70 |
| Byron Buxton | 2016 | BA | 70 |
| Carlos Correa | 2015 | MLB | 70 |
| Corey Seager | 2016 | MLB | 70 |
| Dustin Ackley | 2011 | BA | 70 |
| Eric Hosmer | 2011 | BA | 70 |
| Francisco Lindor | 2013 | MLB | 70 |
| Jesus Montero | 2011 | BA | 70 |
| Jurickson Profar | 2013 | BA | 70 |
| Lonnie Chisenhall | 2011 | BA | 70 |
| Manny Machado | 2011 | BA | 70 |
| Mike Trout | 2012 | BA | 70 |
| Nick Castellanos | 2013 | MLB | 70 |
| Nick Swisher | 2002 | MLB Scouting Report | 70 |
| Nick Williams | 2015 | 2080 | 70 |
| Ozzie Albies | 2017 | BA | 70 |
| Rougned Odor | 2014 | BA | 70 |
| Stephen Drew | 2004 | MLB Scouting Reports | 70 |
All these players were expected to produce a .300 or higher batting average with the high Hit grade. Some have. Some haven’t had the opportunity yet. And for those who have failed, here is why.
- Buxton: Contact issues.
- Ackley: Not enough power behind the contact, low BABIP (.279 for career).
- Hosmer: Not a complete failure but swings out of the zone too much leading to weak contact.
- Montero: No plate discipline.
- Chisenhall: Swings too much out of the zone.
- Castellanos: Swings at nonstrikes and has contact issues.
- Odor: Swings at everything.
- Drew: He may have disappointed but performed as expected until falling apart after the 2010 season.
The one common denominator holding most of these guys back from reaching a high batting average is plate discipline with a couple having contact issues (from probably swinging at pitches out of the zone).
With some idea of what the Hit grades are missing, I will move onto the common traits of good MLB hitters.
Determining MLB hit tool
This process is fairly straight forward, find the features of good MLB hitters as a reference to use with prospects. First, here are the top and bottom 10 hitters since 2008 ranked by Contact% (min 2000 PA). The ability to make contact is one of the main traits included in the Hit tool.
| Name | Contact% |
|---|---|
| Marco Scutaro | 94.5% |
| Juan Pierre | 93.9% |
| Jeff Keppinger | 92.7% |
| Placido Polanco | 92.3% |
| Ben Revere | 91.7% |
| Michael Brantley | 91.4% |
| Alberto Callaspo | 90.9% |
| Denard Span | 90.7% |
| Jamey Carroll | 90.5% |
| Nori Aoki | 90.4% |
| Adam Dunn | 70.4% |
| Pedro Alvarez | 69.7% |
| Jarrod Saltalamacchia | 69.3% |
| Carlos Pena | 68.6% |
| Giancarlo Stanton | 67.8% |
| Ryan Howard | 67.7% |
| Chris Davis | 67.7% |
| Miguel Olivo | 66.3% |
| Mark Reynolds | 66.2% |
| Chris Carter | 64.9% |
The biggest take away from the above list is the division of slap (high contact) and power (low contact) hitters.
The next step is the part of this procedure I have the least confidence in. How to split the power aspect between the power grade and the hit tool. Sadly, I agree with Kiley when he says the Hit tool must contain “…bat speed, raw strength and the basic structure of the swing.” Some line drive power aspect belongs to the hit tool, to get hits, but separating it from the Power grade is where I don’t feel perfectly comfortable.
I have an answer but I may easily change my mind in the future. I went with: All hits in play minus infield hits (Speed component) divided by balls in play minus infield hits. A modified BABIP. The extra base hit components (~ISO) will be the Power grade. Basically, can the batter hit the ball hard enough to get on base? It’s probably not perfect but it will allow this study to continue. Feel free to insert your own opinion on this matter. Here the top and bottom players for this metric.
| Name | Hit% |
|---|---|
| Joey Votto | 39.1% |
| Paul Goldschmidt | 38.2% |
| Miguel Cabrera | 38.2% |
| Mike Trout | 37.5% |
| Giancarlo Stanton | 37.5% |
| Chris Davis | 37.2% |
| Freddie Freeman | 36.7% |
| Carlos Gonzalez | 36.3% |
| J.D. Martinez | 36.3% |
| Chris Johnson | 36.0% |
| Zack Cozart | 26.9% |
| Alcides Escobar | 26.8% |
| Casey Kotchman | 26.8% |
| Juan Pierre | 26.8% |
| Nori Aoki | 26.2% |
| Andrelton Simmons | 25.9% |
| Yuniesky Betancourt | 25.7% |
| Darwin Barney | 25.5% |
| Jeff Mathis | 25.3% |
| Brendan Ryan | 24.8% |
The top table contains better hitters but some lack plate discipline. As we saw with the failed prospects, most of these guys were swinging at pitches out of the strike zone. Here are the top and bottom hitters who chased too many pitches.
| Name | O-Swing% |
|---|---|
| Daric Barton | 16.1% |
| Bobby Abreu | 18.1% |
| Marco Scutaro | 18.1% |
| Josh Willingham | 18.6% |
| Chris Iannetta | 19.3% |
| Carlos Santana | 19.6% |
| Jamey Carroll | 19.6% |
| Chone Figgins | 20.0% |
| Geovany Soto | 20.1% |
| Kosuke Fukudome | 20.7% |
| Josh Hamilton | 39.3% |
| Alfonso Soriano | 40.6% |
| Delmon Young | 41.2% |
| Jeff Francoeur | 41.4% |
| Adam Jones | 41.6% |
| Salvador Perez | 42.0% |
| A.J. Pierzynski | 43.3% |
| Miguel Olivo | 43.4% |
| Pablo Sandoval | 44.2% |
| Vladimir Guerrero | 45.5% |
Having a great strike zone understanding and not swinging at pitches out of the zone doesn’t make a hitter good. But on the other end of the spectrum, not being able to identify pitches out of the strike zone can limit a hitter’s potential.
Now I get to the hardest step, how to combine the three sort-of-related factors into a single hit factor. I messed around with different combinations and came up with the following simple equation and subsequent leaderboard:
2 * Hit Rate – O-Swing% + Contact% = MLB Hit Tool
| Name | MLB Hit Tool |
|---|---|
| Marco Scutaro | 134.7% |
| Joey Votto | 133.7% |
| Mike Trout | 132.4% |
| Joe Mauer | 131.6% |
| Matt Carpenter | 131.4% |
| Paul Goldschmidt | 130.5% |
| Michael Brantley | 130.3% |
| Kevin Youkilis | 130.1% |
| Bobby Abreu | 129.5% |
| Jamey Carroll | 129.0% |
| Vladimir Guerrero | 100.4% |
| Brendan Ryan | 100.4% |
| Delmon Young | 100.1% |
| Mark Trumbo | 99.6% |
| Danny Espinosa | 98.6% |
| Alfonso Soriano | 97.8% |
| Adam Jones | 96.7% |
| Jeff Francoeur | 95.9% |
| Jeff Mathis | 89.1% |
| Miguel Olivo | 86.5% |
Looking over the list as an idiot test, three names may stick out as out of place, Scutaro, Carroll, and Guerrero. With Carroll and Scutaro, people must remember this tool is just for hitting and getting on base (~.350 OBP for the pair’s career). They mainly had gap power which puts them above the like of Denard Span who uses his Speed to get on. As for Guerrero, he was at the end of this career and was really struggling at the plate with less power.
Now, I have a good idea of the components for a good MLB hitter. I am ready to move onto the final step, comparing the three above components to the hit grades of recently promoted prospects to see if they match.
Comparing Hit grades to results
For this section, I took 25 prospects (almost equally distributed between Baseball America and MLB.com) from the past three seasons who debuted and had 200 plate appearances. Here is how their Hit Grade compared to their results.
I expected the correlation to be low because the grades are at five point intervals, but not this low. At least the Contact% and O-swing% are positive. The Hit% is a negative correlation. Yep, the higher the Hit tool, the fewer times a prospect gets a hit when he puts it in play. This is an issue.
Now, here is the combined equation compared to the Hit grade:
An r-squared of 0.01? A negative correlation? Basically, the Hit tool is a useless component to determine hitter value as it’s currently being distributed. It shows a small ability for making contact and plate discipline but just barely. Additionally, it misses on a hitter’s ability to generate hits.
I think it’s time to see if a better solution exists in correctly predicting the Hit tool. I have some ideas but those will have to wait for a future article. Until then, happy prospecting.
Jeff, one of the authors of the fantasy baseball guide,The Process, writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won four FSWA Awards including on for his Mining the News series. He's won Tout Wars three times, LABR twice, and got his first NFBC Main Event win in 2021. Follow him on Twitter @jeffwzimmerman.
“Basically, the Hit tool is a useless component to determine hitter value as it’s currently being distributed.”
This is kind of a big deal.
Joey Gallo gives this statement a thumbs-up.
Gallo with two straight 50’s from BA (2015 and 2016). MLB.com went below average with 45, 45, and 40 (2014 to 2016).
How was he ever in the ballpark of 50? They can raise it as he learns how to hit. I don’t get why it gets so inflated. It is just a current assessment of his current skills.
Thanks for this Jeff, I was one of the complainers on the 1st post. This is pretty eye opening even at low N (25). It really suggests that scouts have no idea how to evaluate the hit tool, which is stunning to me. What happens if you take highest minor league level % and correlate with 1st year MLB %s?
The 25 sample stood out and I have gone back and added 14 more from just the past 3 season (all MLB.com). The correlations are even flatter with r-squares of 0.00, 0.02, and 0.07
As for high minors, I don’t have Contact% and O-swing% yet but I am working on it. One study I want to do is the chance hitters improve discipline skills from the low to high minors. Also, Contact and Hit rates. I think way too much projection is baked into the grades and people anchor themselves too much to the initial grades. The Hit tool grade should be in flux but decreasingly so as the hitter approached the majors.
It’s always been frustrating for me that ‘plate discipline’ is lumped in with ‘hit tool’. They seem like very different, discrete skills!
Agreed. Why not make them “Swing Tool” and “Eye Tool” or something. You could average them if you wanted to make the “Hit Tool,” but I think that would properly penalize folks that have no discipline and end up being terrible hitters because of it.
How does that work for a Vlad Guerrero type? 80 hit / 20 discipline – 50 hit tool?
Agreed, but they are intertwined. Bad plate discipline can completely ruin a hit tool, or it can help it greatly. IMO they are very closely related. Sloppy swings and no plan, which is lack of hit tool, leads straight to poor plate discipline as you K a lot. Imagine if a guy could really hit off a tee (see all of them), but couldn’t hit live pitching because he didn’t recognize anything or had a swing that only works in fantasy world (BP, etc). I think the evaluators role is to determine how that swing will work against advanced competition – hence the tough job! In the end it is just a guess – that is the most important thing to keep in mind.
This is very interesting work Jeff, but there is always a part of me that wonders why so many analysts seem to think it’s necessary to attempt to reduce everything to a single number.
Even if you had produced more satisfying results, would we really gain more insight by trying to jam three already valuable metrics (plate discipline, contact, and power) into one number?
It’s not clear to me that any single “hit tool” metric is going to tell us anything more than what can already be gleaned by looking at a player’s bb%, k% and iso in tandem.
Thanks. I think the hit tool should be several components and not so focused on batting average.
K%-BB% and AVG exit Velo, seems like a good way to go back and look at data in order to put a grade on what has already happened. Exit Velo is a good measure of ball striking ability, and K% is a pretty solid predictor of success. Mix in the BB% and you would have a nice picture of ability with the bat.
Of course, that would be impossible to do if you’re just putting a grade on a High Schooler you’ve seen twice.
Agree it is hard with a high school kid. You make a call and give him a 70 but within a year hopefully the kid has 500 PA against drafted talent to judge him.
I know Buxton missed some time but showing a horrible eye and contact issues, he still got insane grades:
Year: Grade by Source
2013: 60 by MLB
2014: 70 by MLB
2015: 70 by MLB
2016: 65 by MLB
2014: 75 by BA
2015: 70 by BA
2016: 70 by BA
In his case, I think people saw too much of his speed contributing to his AVG. Speed is a separate component.
Excellent summary of WAR! A single number is great if you can’t follow the larger conversation or draw your own conclusions.
Ugh, seeing that 75-grade on Taveras still hurts. I really believed he was gonna be a superstar. Oh what could have been. Hope he rests in peace.
I was expecting to see the following:
Buxton: Contact issues.
Montero: no plate discipline.
Tavares: dead.
Fortunately Jeff has far more class than me.
I’ve been thinking about how frustrating hit tool grades are lately, thanks for continuing to get the conversation started. I recently read through old BA lists, glancing at the names and grades and its incredible how wrong some of the grades are. Sure, prospects bust a lot, but it’s still insane how many 70 grade hit tool guys can’t hit. I would love to see a system for breaking it into multiple parts.
Any hit metric that ends up with Joey Votto and Mike Trout on top is along the right path, for sure.
Good work!
Thanks
It feels like power is weighed too much in hit% just looking over that list a lot of high k guys getting a big bonus there. And if we’re going to count power as a part of the hit tool even though it’s got a separate one we should count speed as well.
Which power guys in the overall value seem out of place? All have a great OBP.
I agree, the Hit% list has too many swing-and-miss guys.
Oh i didn’t mean in the overall. I just meant that hit% contained too much power by which I really just meant swing-and-miss and that likely gave a boost to some guys who might not have deserved it on the overall, even if they didn’t show up at the very top. This also might just be me, but I’ve always associated the plate discipline part of hit tool as more not swinging at bad pitches rather than getting a lot of walks a la Mookie Betts getting a plate discipline boost instead of a guy who gets pitched around. Very interesting article though, thanks for the response as well.
If I were you – I’d try to factor in strikeout rate. Vlad rarely struck out which is why he was a valuable hitter despite swinging at everything.
I wonder how predictive prospect grades are of playing time. Do prospects with higher grades see more opportunity and have a longer leash than those with lower grades? Are prospects with better numbers at a given level passed over for a promotion in favor of other prospects with worse statistical output but with a better grade from scouts on their hit tool?
This certainly happens. A year or two ago, the Rays had some guy named Joey Butler who appeared out of nowhere and hit well for a while. Go look at his FG page and check out his minor league numbers. That guy was a monster.
But he didn’t get any real run until 2015, at which point he was 29, I think. I guess to be fair I don’t know any of his scouting grades either, but I’d be willing to bet they weren’t very good.
Anyway, I think most teams probably favor scouting grades to minor league stats. I’m sure you could find tons of examples of guys with awesome MiLB numbers that never got an MLB opportunity.
I’d be curious to see how analytically inclined clubs are evaluating the ‘hit’ tool. I’m willing to guess that the A’s are probably not weighing the hit grade very heavily or maybe using it with normalized stats for league difficulty when evaluating potential hit tools.
I have long thought that the way hit tool is graded is flawed. Too many guys are within the 40 – 60 range. In reality there should be lots of 30s and lower in this high-K, flyball era. I just don’t think scouts are great at evaluating it as it is all judgement. I see pull-heavy sloppy swings passed off as beautiful all the time. When one scout says it, next thing they are all saying it. Judgement is tough.
It is a lot easier to look at measurables. I can assure you that no computer-based metric will succeed either – exit velocity would be the worst tool to use. I wouldn’t be surprised if hit tool was easier to evaluate without the noise which advanced metrics provide.
You have to know what a balanced swing and approach looks like. Not many people are suited to the task. For a recent example, how do you quantify what Mickey Moniak has? That is a guy with a 70+ hit tool. It is no guarantee of success, but it is something that a trained eye can see.
This is really excellent work to see what a lot of us have always assumed play out in numerical form. I can’t say I’m surprised to see a results-based orientation produce such scattershot findings. I’d be curious to see results based on things like line drive rate and maximized launch angle to see if hit-type metrics like those produce similarly dubious correlations. Interesting stuff as always, my dude.
Should look into potential survivor bias here. Players that perform below a certain level won’t make it to the PA threshold, which would flatten the relationship if it exists.
Could the source have something to do with it? Perhaps MLB.com and Baseball America have diverging views of the Hit Tool. It seems that the 60-grade hit tool was favored overmuch – does this suggest a bias in the scouting reports?
Are we overthinking this? I always thought of the hit tool as measuring OBP, and power tool as measuring ISO. If you want to dive into their components (launch angle, exit velocity, various contact/swing %s), you can, but I’m not sure you’d glean much add’l info beyond their actual results, since this is a retrospective exercise, not a future projection.
Might see a stronger correlation with sources other than BA, too, though it’ll be a few years before Kiley’s earliest ratings can be assessed in comparison.
Cool stuff. I’ve long been skeptical of the scouting hitting grades – though I don’t know how much of that to attribute to trauma from Ackley and Jesus Montero’s adventures in disappointment.
I am curious, though, about how well your new metric (which anecdotally makes a lot of sense) correlates to MLB batting average. Do you have that information?
What if you try to separate out the two publications when you test for correlations? As in, who ever said that an mlb.com 70 is anywhere near the same as a BA 70? I mean unless of course all 25 prospects in question got exactly the same grade, but it seems to me this is a fairly simple variable to eliminate.