In baseball, prospects are generally seen as a crapshoot, and it is easy to see why.
According to Scott McKinney of Royals Review, of Baseball America’s top 100 prospects from 1996-2003, nearly 70% of them posted Wins Above Replacements (WAR) on average of 1.50 or less. Considering that these are supposed to be the premier young players in the sports, such a high rate is alarming.
However, what if this has a way to do with how prospects are evaluated? Currently, there is a lot of focus on subjective evaluation. Whether it be having scouts go to games, watching them take batting practice, or the overall reputation of the prospect, a lot goes into judging them. Perhaps, if we looked at them similar to current major leaguers, our ability to evaluate them properly would improve.
This line of thinking inspired to analyze the stability of these prospects’ statistics. Our ability to evaluate major-league players is due in large part because their numbers are the driving force; we look at them much more objectively. Is the same true with amateur players? If so, does this mean that there is an objective way to analyze prospects, which can increase our success rate in projecting their major-league production? Without further adieu, let us dive into the numbers!
College To Minors Hitters
For a prospect to be acquired by an organization, they are either acquired through the draft or as an international signing.
Low-level minor-league statistics have been cited as being relatively insignificant, and with there being no statistical knowledge of a teenage prospect, they are “toolsy” projections. Hence, why they are more volatile.
Players drafted from college, however, do have public statistics attached to their resume. Yet, are these statistics relevant when projecting prospects to the next level? To answer this, I examined college players that have been drafted from 2016-2018. This sample size was used in order to analyze prospects’ immediate transition to the minor leagues, as performing well immediately is useful for the following reasons:
- More likely to lead to a smooth progression curve to the majors
- Avoids statistics being less legitimate due to being older than the competition or repeating levels for multiple seasons
So, is there a clear correlation between college statistics and minor-league statistics? After further review, it all comes down to using the right statistics:
College To Minors Batting Average
College To Minors On-Base Percentage
College To Minors Walk Rate
College To Minors Strikeout Rate
College To Minors Isolated Power
Based on the coefficients of determination (r^2) for the relationship between college-to-minor transitions for each of the major five offensive statistics, we see a predictable trend. As discovered when analyzing the stable metrics for MLB players, power production and getting on base are far more stable than hitting for average, while plate discipline statistics are the most stable. Here, current strikeout rate remains incredibly predictive of future strikeout rate, while batting average is by far the least stable statistic.
What is interesting is that walk rate isn’t as stable as one may have though, although the r^2 jumps up in a notable fashion (37%) should you take out Seth Beer, who is a clear outlier; his college walk rate of 21% has dropped to around 8% in the minors, which is certainly unprecedented. There are many reasons why this could be true. Although walks are a demonstration of plate discipline, they also are a reflection of how the opposing pitcher sees you as a threat. Players drafted in the top five rounds are generally the top players on their respective teams, making it more likely they would be pitched around. Thus, the transition from college to the minors sorts out those who legitimately have strong plate discipline from those who may have artificial walk rates. Regardless, it remains a better assessment of on-base ability than batting average, which is reliant on a number of factors, many of which center around luck.
Although strikeout rate is rather stable, it is also the one statistic listed that doesn’t have a direct relationship with production. For this reason, it remains the least predictive of a college’s players minor-league numbers, meaning that strikeout rate isn’t a major detriment for a college player; it is a red flag if the whiff at a high rate, but one can still hit for a high average and get on base in spite of it.
These correlations were stemmed from players at Power-5 conference, but when we extend this to players from non Power-5 conferences, the statistics are less reliable. Now, plate discipline rates maintain their stability, but batting average and on-base percentage see their r^2 decrease by 7%. The main difference, however, is power production. The r^2 for isolated power from college to the minors when including non Power-5 conferences drops from 40% to 25%. That is a massive amount, which means that a hitter’s peripheral statistics are even more significant for players at non power-five conferences, while the anticipated learning curve is also much steeper.
College To Minors Pitchers
In general, pitchers are more volatile than hitters. However, they are are generally less subject to the level of competition than hitters, who engaged in reactionary actions. With that in mind, the general learning curve for college pitchers is less. However, that doesn’t particularly hide the fact that they remain volatile:
College To Minors ERA
College To Minors WHIP
College To Minors K/9
College To Minors BB/9
College To Minors HR/9
First off, if you are using one’s ERA at the college level, or any surface-level statistic, to judge their overall talent, you’re doing yourself a disservice. ERA is so reliant on one’s defense, ballpark dimensions, and other factors, and generally, is almost always the least stable metric for pitchers regardless of level.
Given what we know about the stable metrics for pitchers, it isn’t shocking that one’s strikeout rate and walk rate are easily the two most sticky statistics from college to the minors. What is, interesting, however, is how much stronger the correlation of one’s walk rate from one level to the next was over their strikeout rate.
Why would this be the case? Once can make the case that walk rate is a more independent statistic than strikeout rate. In other words, it is less affected by the strength of opposing batters than strikeouts, and that truth is even greater when comparing it to any batted-ball statistic- home run rate, for instance.
This is actually good news for evaluating pitchers. Command/control is incredibly subjective, so the fact that there is an objective way of judging it is incredibly useful. Also, there are a lot of instances where a pitcher who didn’t strike out many batters in college had more success than in the minors. This would speak to the untapped potential of pitchers if optimized properly, which means that there may be hidden value in pitchers who can limit walks and don’t have elite strikeout rates in college. Also, although the drop-off is naturally greater for them, pitchers who strike out 10+ batters per nine continue to do so, even if their pitch data isn’t up to par. This is great news for the likes of Reid Detmers, Nick Swiney, and Logan Allen; college pitchers from this past draft who don’t throw super hard, yet, for some reason, have a knack for striking batters out. Thus, pitchers who produce in college may be overly-criticized by the notion that “it won’t translate to the next level”.
We’ve gone over the notion that pitching is much more independent than hitting, which would explain this; the stability of statistics doesn’t decrease when you factor in non Power-5 pitchers. The learning curve for strikeouts is greater for non Power-5 pitchers, but there should be less concern about quality of competition for pitchers than hitters.
Minors To Majors Hitters
Now that we’ve established the fact that the correct statistics are useful for projecting college players to the minors, is the same true from the minors to the majors?
To answer this, let us look at the transition for players from the upper levels to the majors. Our sample size will from 2015-2019. Ideally, the stability of a player’s numbers will get even stronger as more information is gained on them, meaning that we can project their future success in a more objective manner.
Minors To Majors Weighted-Runs-Created-Plus
Minors To Majors Batting Average
Minors To Majors On-Base Percentage
Minors To Majors Walk Rate
Minors To Majors Strikeout Rate
Minors To Majors Isolated Power
As you can see, the statistical correlation of statistics from the minors to the pros is far greater than that from college to the minors.
As expected, it is all about choosing the right statistics. Expecting a batter to repeat their on-base percentage, wrc+, or batting average without ay context is foolish. However, relying on peripheral statistics appears to be incredibly important.
For instance, strikeout rate in the minors is as predictive of one’s major-league on-base percentage as their batting average. Thus, taking both into account when projecting one’s batting average is key. Based on its direct correlation with success and overall statistical stability, meanwhile, walk rate is incredibly important to use when judging minor-league players. By then, we’ve sorted out whose walk rates are a reflection of strong plate discipline from those who were just the best players on their college teams.
Often times, evaluators see walk rate as something that can improve over time. I have often objected, as I see plate discipline as more of an innate ability. This research would appear to back up that notion. Those who don’t walk in the minors don’t tend to walk in the majors, and the same is true vice versa.
Meanwhile, if one’s power potential is still untapped in the upper levels in the minors, there is generally cause for concern. We’ve seen many players with big frames or high exit-velocity numbers underperform expectations due to an unoptimized swing, and at some point, raw ability needs to translate into in-game production. If not, you may catch yourself dreaming on a prospect, rather than setting reasonable expectations.
So, in all, it’s relatively similar to college, except for walk rate takes his place as an incredibly important statistic when predicting a player’s future success in the majors. If you analyze a player by their plate discipline numbers and power, you should theoretically do well when it comes to projecting offensive production.
Minors To Majors Pitching
We saw the stability of key statistics rise significantly when comparing players from the minors to the majors. Is the same true with pitchers? Also, do strikeout rates stabilize by then? Let us take a closer look.
Minors To Majors xFIP
Minors To Majors ERA
Minors To Majors FIP
Minors To Majors K/9
Minors To Majors BB/9
Minors To Majors HR/9
Once again, using ERA to judge a pitcher’s talent level is not a good idea. Luckily, if you’re looking for one “tell-it-all” statistics, fielding independent pitching (FIP) and xFIP, which regresses home run rates, are pretty reliable.
Interestingly, the stability of K/9 and BB/9 flip here. This may have something to do with the quality of competition stabilizing in the upper levels of the minors. Once again, this is a strong notion in support for those who overachieve expectations by striking batters out without flashy velocity. Meanwhile, it is unclear why walk rate becomes far less predictive, but this is all in line of what you’d expect based on year-to-year correlations from the majors.
One strikeout is much more important than one walk prevented. However, when considering the extra boost in stability with strikeout numbers, finding pitchers who can miss bats remains the ultimate goal. By now, pitchers should theoretically have been developed and optimized, meaning that their potential should line up with their production. If not, then that is a major concern. Two examples of this would be Sixto Sanchez and Dustin May. Neither missed many bats in the minors, and that plagued them in their rookie seasons. It will be interesting to see if their high velocities translate into more strikeouts in the future, or their lack of elite minor-league production should have been a greater concern. I’d side with the latter, given how close the minors-to-pro correlation line up with the year-to-year stability of statistics at the major-league level.
So, are amateur baseball player statistics useful when judging the talent level of a player?
In my opinion, this is undeniably the case. The correlation of minors-to-majors success lines up with the year-to-year correlation of success in the majors. Meanwhile, there is more variance with college players, but not enough so to believe that statistics shouldn’t be the driving force of their evaluations.
Now, you have to look at the proper statistics. For hitters, that means plate discipline and power. For pitchers, that means their K-BB ratios. Simply expecting a player to continue to have success is foolish. Rather, how they come about that success is more useful when projecting them to the next level, which is why peripherals are so useful.
Also, context plays a major role. The learning curve for non Power-5 hitters from college to minors is far steeper than that of Power-5 hitters, and since we were only looking at players who debuted in the majors from ages 19 to 25, we didn’t have to worry about players being older than their competition. Adjusting statistics is incredibly useful. If not, then your projections will be more faulty. Pitchers, on the other hand, are much more independent, so quality of competition is less of a concern.
Hopefully, if the right statistics continue to be more of a driving force for prospect evaluation, certain players will no longer be overlooked. There are several examples of players who produced in the minors but didn’t look the part, only to succeed in the minors. Generally, teams don’t look much at media prospect rankings, yet fans do when analyzing their prospect acquisitions, whether via trades or the draft. Thus, providing them with more accurate information would be beneficial. Doing so by relying more on objective evaluation methods would appear to be in order. The more subjectivity there is, the greater room there is for error. With how volatile prospects are, we should continue to strive to decrease that room for error.
Baseball is made overly complex, but as Brad Pitt, playing Billy Beane, says in Moneyball, “If he’s a good hitter, why doesn’t he hit good?”. Raw talent should translate into in-game production at some point. This line of thinking may seem way too simple, yet it remains the most fool-proof way of evaluating prospects. After all, we aren’t searching for Fabio, but rather, productive baseball players!