The whole idea is to strand as few base runners as possible. A team that hits a lot of doubles (i.e. the Doublers) will do a much better job of this than the Piranhas. In practice, it's very difficult to find enough hitters that hit home runs enough to stack a line up, but good doubles hitters are more common. A high walk rate is usually indicative of a patient approach to hitting. Part of the reason we want this is not just because they're getting on base through walks, it means that they're waiting for better pitches to hit. I disagree with having good hitters auto-take a pitch on 3-0 for this reason. If the pitcher is going to throw it right down the middle, you should get ready to whack it. Furthermore, if the pitcher knows that you're sitting on a fastball right down the middle, it will force him to be a little more careful in what he throws and increase the chances that you walk on 3-0.

Moss has a point relating to Part I, also. Does the results of Part I indicate that the Bert Blyleven pitching mantra, i.e., lots of solo homers won’t hurt a pitcher, is flawed? Wouldn’t Part I indicate that a pitcher is OK giving up lots of singles, as long as they don’t allow lots of extra-base hits? This is probably best modeled with extreme GB pitchers, who are normally very successful. Any thoughts??

I think Blyleven's mantra is based on the fact that even if you throw a fastball right down the middle, hitters won't be able to hit solo home runs enough to really hurt you. Obviously though, any time you give up a run as a pitcher it hurts you. I think his main point is to try and stay aggressive and dictate the game rather than to become passive and let the hitters affect you too much. If you consider particular game situations such as 2 out and nobody on base, you can give up a lot of walks to equal the value of a single solo home run.

I would tend to believe that GB pitchers should be better than fly ball pitchers since they will not only give up fewer HRs, but fewer extra base hits as well. Typically, most flyball pitchers have to also be good strikeout pitchers to make up for this.

I was wondering this too. I tried to work on this problem a bit, but haven't come up with a good way to attack it yet. No matter what they do, they can still not win more than 50% of the games (since the sharks distribution is optimal). I tried to do some stuff that gave me an upper bound on the piranhas win percentage, but then decided that I was making unreasonable assumptions. You should be able to compute the probabilities directly but it might become extremely cumbersome. I think you'd need to do simulations to really get an idea of this. My guess is that a dynamic strategy for the Piranhas would win about 47-48% of the time.

Yes, but don't flyball pitchers (generally) have higher strike-out rates?

*goes to work plotting

I think it's a little tough to relate the two situations. From what I can understand of the Blyleven/Schilling school of thought, the idea is that with no one on base, you'd rather challenge a hitter than allow a walk. A side effect of this is that you'll allow more home runs, but the pitcher certainly isn't trying to induce a home run. The choice is between risking contact and risking a walk, which isn't quite the same as choosing between allowing extra base hits and singles/walks--it depends a lot on how much you're increasing the risk of extra base hits and how much you're reducing the chance of a walk. (It also seems like the cost of a walk would be a lot different on a team that has only piranhas compared to the cost of a walk against a team that can hit a home run.) It also assumes that the pitcher has a degree of control over when he will allow home runs and when he will allow walks. If that's true, then it seems unlikely he would risk allowing a home run in a one-run or tied game.

Wouldn’t Part I indicate that a pitcher is OK giving up lots of singles, as long as they don’t allow lots of extra-base hits? This is probably best modeled with extreme GB pitchers, who are normally very successful. Any thoughts??

I do think this study provides a good reason to choose ground ball pitchers over fly ball pitchers--all else being equal.

*a tear rolls down cheek*

As we are clearly seeing, the singles-hitting offense is very poor...and the Twinkies don't hit anywhere near the .360 required to get to 4.5 runs/game with just singles. (Of course, no team, or even player, hits homers at the rate required to get to 4.5/game, so obviously the model is extreme.)

Moss'll second the kudos on the great work, GreekHouse!

Moss has a point relating to Part I, also. Does the results of Part I indicate that the Bert Blyleven pitching mantra, i.e., lots of solo homers won't hurt a pitcher, is flawed? Wouldn't Part I indicate that a pitcher is OK giving up lots of singles, as long as they don't allow lots of extra-base hits? This is probably best modeled with extreme GB pitchers, who are normally very successful. Any thoughts??

So, for the bunting piranhas, it only makes sense that they'd get beaten by everyone else in 9 inning games, since they only score 4.0 R/G. Depending on how you are simulating this, this could be easy or hard to answer, but I'm curious: how good would the bunting piranhas be if they had a dynamic strategy?

From what I recall, playing for one run is best in the late innings (say 7th inning and on) when you are down by one run, tied, or ahead by one run. (From what I recall from Baseball Between The Numbers the last situation--ahead by one run--is actually a really good place to play for one run, which surprised me somewhat.) So I would be curious to see how much bunting helped the piranhas if they did it situationally, which is probably what gives them the best chance to win.