Last week, I examined whether experienced teams win more often in the playoffs, and found that the answer appeared to be that indeed they did. The overwhelming response I got, however, was negative. Take a look, for example, at Mitchel Lichtman’s (MGL) comment on Ballhype:
I’m still not buying your conclusion until I see “why?” At the very least, I want to see how players with lots of experience did in the postseason relative to their regular season as opposed to players with little experience. The assumption (if you buy David’s conclusion) is that players with postseason experience will “outplay” their lesser experienced counterparts during he post-season[…]
For example, if it could be shown that players with more experience simply perform better than expected in the postseason, then our inquiry is pretty much over. If not, then we’d have to dig deeper on a team level, which has a lot more noise in it. David, you started with something very noisy (team-level win loss records, or series records, which is even more noisy).
A lot of comments I received were similar to this one. Now, I do think that the team numbers might show us some effect that won’t appear in individual statistics, but I will admit that if experienced teams win more often in the postseason, experienced players in all likelihood perform better in the playoffs as well.
We’ll start with hitters, for whom my metric of choice is weighted on-base average (wOBA), a statistic which expresses a player’s total offensive output on a scale that resembles on-base percentage. Here is, since World War II, the average drop-off from a player’s regular season wOBA to his post-season wOBA, based on the number of times that player had previously been in the playoffs:
Experience PA wOBA_Diff 0 22928 -0.033 1 15505 -0.032 2 10402 -0.031 3 6355 -0.036 4 3361 -0.047 5 2083 -0.019 6 1254 -0.026 7 727 -0.016 8 407 0.019 9 179 -0.123 10 76 -0.044 11 47 0.081 12 4 0.071 13 1 -0.361
What this table tells us is that 22,928 plate appearances, players with no prior post-season experience saw a 33 point drop-off in their wOBA in the playoffs. That number holds true for the next few years, but you may notice a sort of inflection point at five years, where the drop-off falls to just 19 points. Here is the data broken out by whether a player had five or more years of playoff experience, or less:
Experience PA wOBA_Diff <5 58551 -0.033 5+ 4778 -0.021
Players with 5+ years of postseason experience see their wOBA drop by 21 points between the regular season and playoffs, whereas players with fewer than five years see a drop-off of 33 points. That difference is statistically significant.
It is also cherry-picked. Since I only picked my cut-off point after running the numbers, traditional concepts of statistical significance go out the window. Instead, all I can truly say about this gap is that it’s interesting. Actually, I’ll say a bit more in a moment, but first let’s look at the pitchers.
Here is a table similar to the one I just presented for hitters, but using runs allowed per nine innings (RA) instead of wOBA.
Experience IP RA_diff 0 6549 0.42 1 3847 0.28 2 2289 0.31 3 1396 0.42 4 821 0.37 5 436 0.10 6 320 -0.71 7 138 0.48 8 150 -0.47 9 90 0.00 10 30 4.68 11 27 -1.63
Here, a negative number is good in that it means that a pitcher allows fewer runs in the playoffs than in the regular season. Pitchers with no playoff experience, for example, see their RA rise 0.42 points in the playoffs. Again, five years of experience looks to be an inflection point, and here’s what happens if we split the players up into two groups: Those with less than five years of playoff experience and those with five or more.
Experience IP RA_diff <5 14902 0.36 5+ 1191 -0.08
Again, there is a significant gap between the two groups—whereas pitchers with less than five years of playoff experience see their RA rise by 0.36 points, those with five or more years actually see their RA decrease by 0.08. In all, that’s almost a half-a-run gap.
But once more, the numbers are cherry-picked, and so there is no statistically valid conclusion to be made. Still, I have to ask, what does this all mean?
It is not, for sure, proof that players with a lot of postseason experience do better in the playoffs (relative to their regular season performance that is) than players who do not have all that much experience. It is, however, evidence.
It might be weak evidence, and for many, it will likely be unconvincing. Certainly, I wonder why it would take five years worth of postseason experience for this effect to show up. However, the effect is there: Players with five or more years of playoff experience perform significantly better in the postseason than their less-experienced counterparts. That is a fact.
It may be cherry-picked, and maybe in your opinion it is worthless, but it’s all I’ve got.