Thursday, January 28, 2010

Diving Deeper into (Patton's) Pitching Valuation

My analysis on A.L. starter vs. reliever ERA/WHIP showed that five of the top 10 ERA/WHIP earners were relievers. BirdWatcher wonders if I'm giving the relievers too much credit.
Is it possible the valuation methodology underlying this discussion overvalues the contribution of relief pitcher ERA and WHIP ? Let's take an extreme case of 65 IP with a 1.10 WHIP and 3.00 ERA and also assume a team total of 1300 IP with a roughly league average 1.35 WHIP and 4.25 era. Absent this reliever, team WHIP would be 1.363 and team ERA would be 4.33. So, the reliever's presence might move you up one standing point each in WHIP and ERA. How much is a one point move worth? Perhaps $1-$2? But the valuations you use are showing figures for ERA and WHIP each in the $4-$5 range.
Let's look at a couple of examples from the American League last year.

In 2009, Mark Buehrle tossed up 213 1/3 IP with a 1.252 WHIP. Phil Hughes threw up 86 IP with a 1.116 WHIP. Both earned $3.15 in A.L. 5x5 leagues last year.

The formula I used for WHIP assumed 1200 IP per team and a 1.360 WHIP. Add Buehrle's WHIP to this team and the team's WHIP improves to a 1.344, Add Hughes' WHIP to this team and the team's WHIP improves to a 1.344.

You can do the same thing for ERA. Jose Mijares put up a 2.34 ERA in 61 2/3 IP. Edwin Jackson had a 3.62 ERA in 214 IP. Both earned $3.10 in A.L. 5x5 last year.

The formula I used for ERA assumed 1200 IP per team and a 4.2275 ERA. Add Jackson's ERA to this team and the team's ERA improves to 4.135. Add Mijares' ERA to this team and the team's ERA improves to a 4.135.

It is important to note that the 1200 IP baseline assumes eight pitchers and that you are calculating value by taking a pitcher and making him the ninth pitcher on this hypothetical team.

You are most certainly welcome to argue about how much each pitcher's contributions are worth. If you prefer to de-value each pitcher's ERA/WHIP contributions, you could push the IP baseline up to, say, 1300 IP. This would make Jackson's ERA worth $2.89 and Mijares worth $2.87. Adding more innings to the formula would make starters more valuable, but only slightly so.

You could also decide that your league had a better ERA/WHIP baseline than the 4.23 and 1.360 that I'm using. In this case, the reliever ERAs and WHIPs would be worth less but so would the starting pitchers'...and by an equal amount.

Put simply, if you want to tamper with the Patton formulas so that relievers ERA and WHIP are worth less, you are certainly entitled to do so. However, starters' ERA and WHIP will also lose value as well. The value of Mijares' ERA doesn't equal the value of Jackson's ERA because Mijares' innings are too highly valued, but because both pitchers had the same impact on their teams ERA in 2009.

2 comments:

Anonymous said...

MRs need to be a important part of any staff. They've saved me year after year.

Eugene Freedman said...

The real problem with middle relievers is that because they throw so few innings the sample size is very small. That makes their ERA much more variable and subject to the effects of one or two bad outings. Therefore, they are much less predictable.

Wins are normally very luck oriented. Pitchers on the same team may have quite disparate run support. But, with middle relievers, usage patterns matter more than anything. If used behind by a run or two it's better than being used to protect a two+ run lead, as counter intuitive as it sounds. The pitcher who the manager relies up less will get more wins.

Projections should consider those factors.