"Moneyball" and Paintball
What teams and players stand to learn from baseball, Bill James and the sabermetrics movement in sports
Tournament paintball has long been considered an intangibles game- to quote Matty Marshall, you want players with “gumption”, players who are “always 8’s, not ‘sometimes 10’s’ and ‘sometimes 2’s’”. You can’t quite tell what makes a player good... hell, you may even know they’re “good at laning” or that they “just shoot people”, but what does that really mean? Can we distill it down any further? In what context do these qualities exist?
In 1977, a Vietnam war veteran named Bill James released an 80-page doctrine on baseball stastics modestly titled The Bill James Baseball Abstract. Loaded with intricate statistical analysis, the book was heralded for both its novelty and complexity but considered to have very little practical use. Baseball was, after all, a sport about sweet swings and hitting homers, not walk ratios and on base percentage. The most important statistic was intangibility- whether or not a guy just had it.
Player evaluation followed this model for more than a century. A square jaw and firm handshake were almost as important as whether a kid could throw or catch. Never mind if he had a high strike-out rate or couldn’t hit a curveball- if he had “it”, he had a contract. Managers just knew.
The funny thing about just knowing, though, is that it fails to produce wins at an alarming rate. Teams like the New York Yankees were doling out hundreds of millions of dollars more than most other clubs and consistently coming up short. The game was practically rigged in their favor- baseball has no salary cap and very loose free agency rules; a fat payroll usually means you can have your pick of the litter as far as talent is concerned. The list of World Series Champions from 1990 to 2000 should have looked as such:
1990: New York Yankees
1991: New York Yankees
1992: New York Yankees
1993: New York Yankees
1994: New York Yankees
1995: New York Yankees
1996: New York Yankees
1997: New York Yankees
1998: New York Yankees
1999: New York Yankees
2000: New York Yankees
That fact that it didn’t pan out that way meant one of two things: either the Yankees’ management didn’t know “it” when they saw it, or “it” didn’t really matter that much in the first place.
Oakland A’s GM Billy Beane believed in the latter. In 2002, using Bill James’ concept that statistics held the key to winning baseball games, Beane assembled a playoff team on less than $40 million dollars (well below the MLB poverty line). Two years later, the Red Sox won it all using James’ formula. (Both teams were featured prominently in Michael Lewis’ Moneyball, the book-turned-movie that chronicles baseball’s statistical revolution.)
The Moneyball method works because it asks a few key questions: what is a win made of? Which numbers matter and which are irrelevant? What is the most efficient way to produce those numbers? In baseball, it was stats like OBP (on base percentage, or how often a player reached base safely) and wins-over-replacement that mattered. Those numbers could be distilled down into wins.
The question is, how can we apply meaningful statistical analysis to tournament paintball? Can we do it at all?
(Not to spoil the ending for you, but I don’t have any answers yet. I’m not even certain metrics hold any value at all in a game with as many variables as paintball. All I know is that it works in lots of other complex sports and is worth trying to apply to paintball in logical and reasonable way)
The following statistics are where I believe we should start. There’s nothing absurd like “first ball trajectory” or impossible to track like “amount of paint spilled on reload”; I don’t believe the answer would lie there even if we could log those things. If statistics do provide insight as to what a win in paintball is made of, I think the big picture ideas will be the real players:
(note: for the sake of continuity, I use the words “points” to describe individual paintball games, i.e. from the start horn to the flag hang. I understand “points” is ubiquitous with XBall, but these statistics should be applicable in the 3/5/7 man formats as well)
1) Living Off Break percentage (LOB): The percentage of points a player starts in which he reaches his primary without being eliminated on the way or in the five seconds that follow.
Why does this matter? Because it tells you which players contribute to your team’s...
2) Wins with an Advantage in Bodies after the Break (WABB): The percentage of points a team wins when up bodies after the break.
This will prove to be an incredibly meaningful statistic if it turns out teams win something like 60% of all points in which they have at least a one body advantage after the breakout. It could even be broken down further into WABB+1 (wins when up one body), WABB+2 and so forth.
Furthermore, if it turns out that WABB is statistically significant, it will also show that winning when you’re down on bodies after the break is more an aberration than a rule and shouldn’t be read into too deeply.
3) Wins-to-Break Shooters: The number in games won with varying numbers of guns up off the break.
Say you know you win 66% of games in which you have four guns up off the break but you only win 49% when you send three people sprinting to the far bunkers with their heads down. You can dial in exactly how many break shooters you should have in a number of different scenarios. You can also use this number to extrapolate how effective each player is as a break shooter- if you have a higher win percentage with four guns than you do five, the variable player may be an ineffective break shooter.
4) Shots-to-Wins Ratio: The number of points a team win when it fires more paintballs than the opposing team.
The NPPL sort of tried something like this in DC- after every pro point the commentator announced who shot the most paint. It was definitely cool to hear, but it doesn’t mean anything unless you correlate it to wins. We obviously have the technology for this one, and it should theoretically be one of the easier stats to track.
5) Live Body Percentage (LBP): The number of bodies the winning team has left after the flag hang.
Tells us how effective a specific team was in terms of both shooting the opposition and not getting shot.
6)Penalties Per Game (PPG): The number of penalties each team gets in a game.
If teams with more penalties consistently win games, it tells us that penalties are statistically irrelevant. If teams with more penalties consistently lose, it tells us that penalties have a major impact.
Couple that with...
7) Penalties per Point Percentage (PPP): the number of penalties a player gets over the number of total points he plays (the higher the percentage, the worse- more penalties per point)
Voila. Now you know precisely who is hurting your team.
I think these seven statistics are a good jumping off point. I don’t believe statistics like “player by player kill count” are either feasible to track or all that relevant- getting the kill is often the easiest part of the job. Somebody else could have been pinching that player out into your lane or drawing their attention or have forced them to move into a less advantageous spot. Overall team kills and difference in live bodies seem (at least on the surface) to be more relevant stats.
But I don’t know that. We don’t know that. If we can extrapolate what matters most in winning paintball games, then we can debunk all sorts of paintball dogma that is espoused but has never been fact checked. What if it’s not better to be short? What if it doesn’t matter? What if it’s better to be tall? What if breakshooting doesn’t matter? What if it’s the only thing that matters and taking ground is an exercise in futility?
These things need to be tracked because the key to consistently winning could very well lie within. The parody at the national tournament level alone should tell us that eyeballing players doesn’t work.
Or maybe it does, and paintball just has an unquantifiable randomness to it.
We don’t know.
We should find out.