Unicorn Hunting - Team and Player 'Runs' values distributions

I decided i wanted to know how teams build their rosters, and how players build their value. I have previously put together a post looking at the distribution of run values for Batting, Fielding, and BaseRunning. Batting runs have a much greater variance than Fielding or BaseRunning runs. As such, we can expect that on average more players will derived more of their value from batting than the non-batting runs. But not all runs created are batting runs, and players have strengths and weaknesses. This gives Teams the opportunities to determine which traits they want to recruit more. In this post, i want to look at how players generate their Runs, and also look at how teams differ in their productivity in aggregate.


I am going to start with teams. i took our five years of data (2018-2023, no 2019), and aggregated all the Runs metrics - BaseRunning, Fielding, and Batting - for each team for all player seasons with at least 200 plate attempts - same data as before.


The table columns here represent each team, their aggregate Batting, Fielding, and BaseRunning Runs, and then total. Recall that an average player-season will have a value of zero runs. So positive runs represent a better than average, and negative runs a worse than average. You will see an imbalance here - teams in this presentation are generally better than average. This derives from the 200 PA threshold. Generally if a player is getting > 200 PA in a season, that player is more likely to be better than average than worse.

The table is sorted by the total runs metric. you can see at the top of the list the juggernauts. Houston in the AL and the Dodgers in the NL are clearly in their own sphere. Atlanta follows not too far behind, followed by the Rays, and our very own Cardinals, who surpass the Evil Empires of the Red Sox and Yankees (yes, I lump them).

The Cardinals rank fifth on this list - which is fantastic, especially given that 2023 happened. That said, their aggregate runs above average is only half that of Houston and LA. This is driven almost exclusively by Batting runs. I will also note that of all the five highest teams, there is precisely one negative value, with Houston having given up about 40 runs (4 Wins) from baserunning over a five year period. Have the Astros decided to ignore baserunning runs and rather draft/sign at the expense of batting? Maybe? Hard do know whether this is a deliberate decision, or a by-product of a 'bring in the best talent' approach. Whatever it is, it has worked remarkably well. I will also note that Houston's Fielding over this time frame is one of the best in the majors. They basically have owned two of the three Run creation means that batters have control over. And the two the own are the two most meaningful.

A narrative, at least for a time, was how Philly basically sold out to offense. It hasn't worked. They are relatively middle of the pack, on offense. They have succeeded remarkably well at deprioritizing defense though, with a massive -81 Fiedling Runs over the five year period. Oakland outperforms Philly by an impressive 78 runs.

The Cards? well, they are pretty balanced (anyone surprised)? They do well on offense, but many teams do better. They do well on Fielding, but many teams do better. The do well on baserunning, but many teams are better. In aggregate, the Cards are a very good team, because they get contributions from each area.

I also want to point out Arizona - they have been worse that average on offense over this time frame, but well above average in Fielding and BaseRunning, resulting in them being just a bit below middle-of-the-pack overall. They are the highest ranked team with below average Batting Runs. Cleveland also stands out with a relatively low offensive contributions. Two thirds of their RunsAboveAverage come from Fielding and Baserunning - a large imbalance compared to most of the league.

We are not privy to internal discussions of these front offices, unfortunately. We also do not know how many of these results are passive vs actively pursued. This table does reiterate in my mind that we are a fairly fortunate group of fans - our team is generally quite good. It also reinforces that we are in a tier below the juggernauts.


I next want to look at how players build their own value. Each players has a value for each Run metric, and some players are all hit, few players are all not. but if we look at all seasons, how much of each component goes into that player's total WAR? There is a nice plot to visualize this, called a ternary plot. It looks like this:


In this plot, we have three 'axes', so to speak, with each represented as the edge of a triangle. Each edge represents a run source, normalized between 0 and 100%. Each point in the plot represents a player-season. Each vertex or corner - the place where two edges meet - gives you an indication how how many players derive their value from the '100%' Run value. For example, essentially no player-season was derived from purely BaseRunning Runs, because the lower right corner is empty, and that is where BaseRunning has a value of 100. You can see more points near the upper vertex, which represents nearly purely Fielding value. Not a lot, but clearly more than BaseRunning. And perhaps unsurprisingly, we see the most points in the lower left corner, representing a high contribution of Batting. There ARE many points which exist outside the corners. These, i would call, more 'balanced' players. They excel in no area, but derive value from multiple.

This plot makes it pretty clear, again, that Batting weighs more heavily, even withing individual player-seasons. Lets go deeper into this. Annother way to look at these data is to calculate the percentage of total runs generated by a player are from Batting runs. To do this, i had to offset all the runs values - adding the minimum run value for all players (which is negative) to every players value, so that all run values are positive. this is shifting the distribution of Runs from centered at zero to centered at the mean runs, and the lowest assigned Run value to zero. Then i added the Batting, Fielding, and Baserunning runs value together for a total. the Batting divided by the total, expressed as a percentage, is what we have at the end of it all. Below, a histogram looking at this distribution:


most player/seasons that have met our filter criterion (200 PA) derive the majority of their value from batting, and it isn't really close. We see a decent peak around 55%. The median value is 53.2%. That is, half the players have a value of greater than 53.2%. The mean is similar, at 53%. The 25% percentile is 46%. 75% of the players generate 46% or more of their value from Batting. The maximum player-season, using this method of estimating percent contribution of batting, is 83%. You can also see that this distribution is skewed right - there are more players above the mean than below, and there are some players toward the left which are FAR below the mean.

- Lets build some rosters from Unicorns.

Rather than just pick the tails of this distribution in prior 'unicorns' posts, I decided i wanted to create teams from it. I divided the players into three groups, based on the percentage of their total Runs from batting:

non-hitters: < 48.6% batting Runs

balanced players: 48.6 - 57.5% batting Runs

hitters: > 57.5% batting Runs

I then constructed the best team i could from each group, to see how these teams compete. I just used WAR to select the best player at each position within each group - no adjustments for plate attemps. I had to pull position data using the player logs for a given season - the position each player is assigned for a given season is the position at which they are listed most frequently. i.e. if Tommy Edman played 100 games at short stop and 30 games at second base, he is listed as a short stop. If he played 60 games at short, and 59 at second base, he is still listed as a short stop.



Our non-hitters lineup yields a total WAR value of 37

We see two Cards on the non-hitters list, Bader's 2018 campaign, and Edman's 2022. A few observations for this table:

1. The Runs here are derived mostly from fielding. Baserunning is about 1/3 the value of Fielding runs, in total.

2. These WAR values are to dream on - to have a team comprised of 3-6 WAR players - well, you are doing quite will, despite a lack of offense.

3. STL is the only team listed twice.

4. Our DH is Byron Buxton ((LINK)), and he was actually worse the average, offensively.

5. 37 total WAR is actually a REALLY good team. This same team, a year later, would be unlikely to replicate that value (we are selecting unicorns, i remind you), but still this is a rather strong team, with less than half the total Runs derived from Batting.



Our balanced player lineup yields a total WAR value of 48

We so zero Cardinals on this list, but one to-be-a-cardinal in Nolan Arenado. This team outperforms the non-hitters by an appreciable 11 WAR. Some observations:

1. More than half of their total Runs are derived from batting.

2. We even now have positive value from our DH - Trey Mancini's 2022 season.

3. Corbin Carroll is the only player who derives more than 7 runs from BaseRunning, and nearly every other player contributes about 1 WAR from Fielding.

4. Nearly all of teh 'percent.batting.runs' values here are from the upper end of our 'balanced' range (48.6 - 57.5%). The lower values on our table come from defense-centric positions - center field and catcher, but even in those cases we see values above 50%.

5. We see Christian Walker here again (2022 vitage) - his 2019 season is also represented for 1B on our non-hitters team.



Our non.hitters lineup yields a total WAR value of 72

This team would crush your team every time. quite literally - i keep hearing the phrase 'Hulk Smash' in my head as i look at it. Pitchers would be cringing in fear, both before the pitch and after the hit, dodging 100+ MPH line drives as if their lives depended on it. Is there anyone on this team you wouldn't want as a Cardinal? Most of them we would still take - years after these prime seasons. Well, all you have is wishing, as there are no Cardinals on this list, or anyone that was a Cardinal prior or in the future. Can you squint at our current squad to see any position player hitting this list in the coming year(s)? I would guess it will take a step up from Gorman or Walker to get the Birds on the Bat onto this list. Observations:

1. this team would not lose many games.

2. this team would be crazy expensive to build.

3. our lowest WAR value here (5.06) is better than 7 of the 9 WAR values from our non-hitters team, and better than 3 of the 9 from our Balanced team.

4. Only one position - Shortstop - derives less than 60% of their Runs from Batting.

5. Four positions (if you include DH as a 'position') - derive > 70% of their Runs from Batting.

6. There aren't actually that many negative values in the Fielding and BaseRunning columns here. While these are members of the 'hitters' group, twe do not see their Fielding and Baserunning as much worse than average, if at all.

Bottom Line(s):

Really this is a long post which tells in more detail us what my prior post did - that Batting is extremely valuable, and in fact more valuable than Fielding and Baserunning combined. Caveat, of course, that this is based on FG Runs and WAR, but i don't think that the overall conclusions would be changed too much by another model.

I always knew that Offense was the far sexier Run creation mode, but I don't think i appreciated how much more valuable it is than Fielding and BaseRunning. That said, it is also clear that there is little dead-weight on the fielding and baserunning values for our 'hitters' unicorns table. They may be bat-first, but they are not all-bat. They still generally derive average to above average value from Fielding and BaseRunning.

This should be the goal for Walker and Gorman. Walker's defense needs to improve appreciably this year - it may, it may not. Here's hoping it does. Gorman is actually holding his own, being only 2 Runs worse than the average second baseman in 2023. Improvement would of course be welcomed, though. They each need to improve their batting by about 10 runs to mingle in the 'hitters' crowd. Is 2024 the year? Fingers crossed.