How do you judge a pitcher?

[Warhammer]

This is coming out of what I think is an interesting discussion in another thread.

How do you judge a pitcher? What stats do you look at?

I look at Ws and ERA for pitchers in combination. Ultimately, the pitcher's job is to put his team in a position to win the game. So a pitcher with more Ws is preferable to one with fewer Ws. ERA to me is an indicator of future performance. It is possible for a pitcher to have a lot of Ws, but a high ERA. To me this is a flag that he is not going to keep this up over time. However, a pitcher with fewer Ws, but a low ERA will probably do better over time. That said, the pitcher with fewer Ws does have some warning signs. Does he give up the big run late, etc.?

Quality stats has always been a made up stat to me. Something to justify why some pitchers are preferable to others. Who cares if you give up 4 runs over 6 innings? That is a pretty high ERA so that means you probably aren't going to win many games. If it was 2 runs over 6 innings, it would mean a lot more to me.

Low BB/9 and high K/9 are nice to have but are not essential. Again, these are indicators. A low BB/9 typically means you are going to have fewer baserunners, but if you give up a couple of walks per game, its no big deal. A high K/9 is great to have, but you could have a ton of Ks, but give up a ton of hits or BBs as well, resulting in few wins and a high ERA.

ERA+ I have always been reticent to use because most people don't know it off the top of their head, but the principle is good. However, I only like to use it if ERA and Ws are comparable.

What do you use to measure quality of the pitcher?

[John Galt]

As I was the other half of the discussion in the other thread, I would refer you to this post:

http://www.operationsports.com/fofc/showpost.php?p=1196416&postcount=54

and this one:

http://www.operationsports.com/fofc/showpost.php?p=1196556&postcount=88

Wins are one of the least valuable stats around (whatever value they have is captured by other stats) and there is little (if any) evidence that some pitchers have the skill of "pitching to win."

edit: BTW, if you are like Warhammer and evaluate pitchers using W's as a key factor, I invite you all to join a fantasy league with me for $100 a person. :)

[SackAttack]

I look at Ws and ERA for pitchers in combination.

Not to ignore the rest of your post, but I wanted to focus on this.

You say that a pitcher's job is to put his team in a position to win, and so W's are an important statistic. I think you're leaving something out.

I think if you're measuring wins and ERA, it's a mistake to ignore no-decisions, because those are ALSO situations where the pitcher possibly put his team in a position to win. Yes, the team may have rallied after his departure, but he could also have left with the lead and had the bullpen blow it, or left in a tie situation.

The reason I think that's important is that the pitcher is still reliant on his offense to score runs in order to get the win, no matter how effectively he pitches. A pitcher with low run support could therefore have a lower win total than somebody else with the same ERA, or even a slightly higher one.

One dude's 3-5 with a 2.90 ERA, the other is 10-5 with a 3.20 ERA. Without looking at the run support situation and the number of no-decisions each pitcher has had that resulted in a win for his team, it's hard to look at pitcher B and say "Well, ERA's only 0.30 higher and he has 7 more wins, he must be better."

[Warhammer]

You say that a pitcher's job is to put his team in a position to win, and so W's are an important statistic. I think you're leaving something out.

Good point, that is something that I would consider, but don't remember the name of the stat. Games won when starting or something was what I remember it being called.

I do understand that run support is critical, but it is something you can't count on. If I lose 2-1, whose fault is that? I would argue that it is mine. The other pitcher pitched better than I did that night. I had a pretty darn good ERA, but I still lost because I wasn't the best pitcher on the field that night. Granted, my hitters are responsible for making the other guy look bad, but still, I have to do what is best in what is under my control.

Now, I win 5-3, does this mean that I played better than the other night? No, but the performance is more valuable because we did win, I didn't blow it and allow the other team to waste our offensive output.

Now, I am not saying that Ws alone determine a great pitcher. It is critical that you include ERA in the discussion for reasons I mentioned. I am trying to find out what other people look at.

For what it is worth, for a closer the most important things for me are BS and Save %. For a relief pitcher it is Hs and fewest inherited runs scored (whatever it is called).

[dawgfan]

Depends on what you're evaluating (and for the sake of this argument, I'm assuming we're talking about starting pitchers).

If I'm evaluating past or current performance, i.e. judging how he's done, I'm mostly concerned with ERA+, with lesser focus on innings pitched and wins. Basically, I'm looking at how well did he prevent the other teams from scoring relative to the league average that season (thus leveling out differences between years and eras) regardless of method - whether he struck out a bunch of guys, got a lot of groundball double-plays, was stingy with walks or was simply the beneficiary of good luck, that's not important to me in this context. John Tudor and Dwight Gooden both had outstanding seasons in 1985; they each did so in different ways, but the bottom line is they both got the job done. I'm looking at innings pitched, because a guy that's dominant over 230 innings is obviously more valuable than a guy that's dominant over 180 innings. I look at wins only as a way to guage how well the team did in that pitcher's starts, but I don't put much weight in this metric since this is so heavily dependent on factors outside the pitcher's control, i.e. how much run support he gets from his teammates.

If I'm comparing similar bottom line results between pitchers, I'll dig deeper and start looking at stats like WHIP and K/9 IP to get an idea of how overpowering a pitcher was, and also to start filtering out the pitchers that had great seasons in large part due to luck.

If I'm attempting to determine how well a pitcher can be expected to perform in the future, I look at a much different set of stats. I look primarily at xFIP, which is expected fielding independent pitching ERA. This metric takes the FIP metric which looks at the factors that a pitcher has the most control over: home runs allowed, walks and hit-by-pitches surrendered and strikeouts to produce a predictive ERA number and normalizes the home-run factor to account for home parks.

This metric is a great way of determining which pitchers are benefitting from good luck or suffering from bad luck. With the research done on DIPS - defense independent pitching stats - we now know that, generally speaking, on the major-league level, most pitchers have no control over whether a ball put in play becomes an out or not. Typically, the BABIP (batting average on balls in play) is around .290. If a pitcher has a much higher or much lower number in this category, this is not a performance that will be sustainable over time. From year to year, this figure will change per pitcher, but for the vast majority of major-league pitchers, they'll normalize over time to around .290 on BABIP.

In addition to xFIP, I'll look at various rate stats and ratios, such as K/9 IP, K:BB ratio, groundball:flyball ratio, HR/9 IP, BABIP. Basically, these are additional means of judging how a pitcher has acheived the results they've gotten so far and evaluating whether that performance is sustainable and matches the expected performance from those peripheral stats.

There are a couple of good examples of the value of these stats on the Mariners. When the M's pursued Jarrod Washburn with big bucks, I was dubious - while he clearly had a good season last year, these advanced pitching metrics showed that he benefited greatly from good luck and that his performance wasn't sustainable. Lo and behold this year, that assessment has proven correct. His peripheral stats haven't changed much - his xFIP is slightly higher this year, but his actual ERA is much higher, mainly because he's regressed to the mean in the number of flyballs that have gone for homefuns, and the amount of runners in scoring position he's allowed to score (he was ridiculously good at stranding baserunners last year).

Another example is Felix Hernandez - he got off to a very hot start last year, and his actual performance closely matched his expected performance. This year his xFIP has gone up a bit, mainly due to a decrease in his groundball percentage, jump in flyballs going for homeruns, and a high BABIP, but the jump in his actual ERA has been much higher than his jump in xFIP. The bottom line being, it's a good bet that Felix, if he keeps pitching exactly as he has, should see his ERA drop as his BABIP normalizes towards the mean. His 4.95 ERA might scare off some casual observers, but he's a good bet to pitch much closer to his 2.67 ERA of last year moving forward than his current 4.95.

[John Galt]

As usual, I think I'm on the same page as dawgfan in a baseball thread. I will say that he doesn't fully address how to deal with the unusual cases (like Webb's groundball rate or Je. Weaver's flyball rate) which change their ability to rely on defense (and who on the defensive side matters more). While he includes those areas generally, there are always some aberrations that warrant further examination (like I've been looking for someone to explain why so many sinkerballers have such extreme inconsistencies over time, but maybe somebody has studied that and I just can't find it).

[ISiddiqui]

Like dawgfan, I start at ERA+. WHIP and strikeout rate (including K:BB) are thrown in for a little bit greater analysis. In evaluating best over history, I'll use ERA+, WHIP, K/9, K/BB and take into account factors like longevity and a pitcher's "best years".

Though, if it is how a pitcher WILL do in the upcoming years, then a whole host of other things come in play... which dawgfan and John Galt go into.

[Grammaticus]

If it is a starter, I look at Innings Pitched, ERA, Opponents batting average, Opponents average in scoring position, BB/9 innings. Then maybe fly ball ratio and HR allowed. I also look at injury history. If they have a propensity to get injured, I usually consider that a pretty big factor.

I would really only factor wins in a lot if the pitcher consistently averages over 15 wins per year. But usually that means the other factors are already really good.

[JHandley]

It seems to me the real crux of the discussion is the level of comparison you're after.

W's and ERA are very simple stats and thus return very broad results. If you just want a general idea of which pitcher is better then a broad difference in W's and ERA will give you an idea of which one is better.

ERA+ and xFIP are more complex stats and thus will return more specific results. It will give you a way to compare the 3-5 with a 2.90 ERA pitcher versus the 10-5 with a 3.20 ERA pitcher.

When you start looking at IP's, OPA, BB/9 and KK/9, that takes more than just an "At-A-Glance" view and of course will give you better results in comparing two pitchers.

[Huckleberry]

If I lose 2-1, whose fault is that? I would argue that it is mine. The other pitcher pitched better than I did that night. I had a pretty darn good ERA, but I still lost because I wasn't the best pitcher on the field that night. Granted, my hitters are responsible for making the other guy look bad, but still, I have to do what is best in what is under my control.

But that's just it, you are not judging them based on what is under their control. You admit that you are judging a pitcher based on how his teammates hit against the opposing pitcher. How is that under his control? Aside from an NL pitcher's theoretical 11% or less contribution to the offense.

[dawgfan]

As usual, I think I'm on the same page as dawgfan in a baseball thread. I will say that he doesn't fully address how to deal with the unusual cases (like Webb's groundball rate or Je. Weaver's flyball rate) which change their ability to rely on defense (and who on the defensive side matters more). While he includes those areas generally, there are always some aberrations that warrant further examination (like I've been looking for someone to explain why so many sinkerballers have such extreme inconsistencies over time, but maybe somebody has studied that and I just can't find it).
Yeah, you can always get more complex in your evaluations. But I think that FIP and xFIP are designed such that whether or not you're a Jered Weaver type extreme flyball pitcher and you have a bunch of DH's in your OF screwing you over, your FIP and xFIP should still show how effective you should be with a normal defense and baseline luck supporting you, even if your actual ERA is suffering because of your defense in the OF.

Also, if you want to dig deeper it's worth charting a guy with a few years under his belt to see whether his BABIP over his career falls close to the expected .290 mark or not. Tom Tippett has shown that some pitchers display some ability to limit BABIP more than the major-league average - Jamie Moyer was one guy listed, knuckleballers in general also tend to do better in this metric, and I think there may be a few other guys that beat the average a bit over time. So while in most cases DIPS theory is correct in saying that a major-league level pitcher has no control over balls in play going for hits or not, in some cases that's not entirely true, and it's worth knowing if you really want to evaluate a pitcher with a significant track record already.

I also haven't fully delved into studies that further break down pitcher tendencies - for example, I'm not sure if there's a baseline level of line drive percentages for pitchers, whether that percentage is related to their groundball/flyball tendencies, and any other possible conclusions from that measurable.

What these metrics also don't account for is the synergy potential between a pitcher and his defense. For example, if a pitcher has an outstanding defensive infield behind him, does that change his pitching patterns? Does that pitcher feature his 2-seamer more to play to the strength of his defense? Unfortunately, I'm not sure this is something that can really be determined with much accuracy from stats given though.

[Peregrine]

I just ask one simple question. Is he a belly itcher? If so, get him out of there.

[Schmidty]

What kind of beer?

[RendeR]

How much beer does it hold?

Is it lightweight and easy to handle?

Can I drink straight from the pitcher without spilling my beer everywhere?

If the answers are acceptable, then I buy the pitcher.

Seems simple enough.


*Please note, acceptability of answers is based ENTIRELY on how much beer has been consumed PRE-questioning....

[Shkspr]

I just ask one simple question. Is he a belly itcher? If so, get him out of there.

Unless, of course, you have a broken ladder at the plate.

[EagleFan]

There is only one way to really judge a pitcher. Watch him pitch.

Watch how he handles batters. Do opposing managers and players change their approach because of him. Does he dominate them? How many broken bats? How many hitters walking back to the dugout staring at the sky in disbelief?

There is no one way to judge a pitcher based on stats. There are way too many variables to the equation. Wins are definitely not the way to judge as that is not something the pitcher always has control over (take some of the Nolan Ryan days in Houston when he was in the league leaders in ERA but had lousy records because he had the worst run support of any pitcher in the league).

Opponents batting average is a decent indicator in some aspects but maybe the average is a little higher because the fielders behind him have limited range compared to other teams. Maybe he gets hit with bad luck bloop singles because he sawed off the batter but the ball just happened to loop in for a hit.

Walks per inning pitched is great for judging control, but that doesn't mean he is a good pitcher. Maybe he throws strikes but gets rocked because of little movement on his pitches.

All in all, there is no one way to judge a pitcher. You need to take all things into consideration.