Historical Pitching Accuracy

StyxNCa · 11-28-2010, 01:22 PM

I simmed the 1969 season 5 times the other day and I was pleased with the overall accuracy of the game. The standings looked pretty good and the individual batter stats looked pretty good. However, the problem I had with the individual pitcher stats is still there for the most part. Many pitchers I looked at had stats that didn't come anywhere close to what they actually did. This has been the problem since the DIPS engine came into play. There is no way pitcher stats will ever look right with this engine.

I am not intending to open an old argument here about DIPS, but saying pitchers don't control the number of doubles and triples they allow makes for some stats that are way off what they actually did.

Just these two examples:

1969 SF pitchers, Gaylord Perry and Juan Marichal.

Real life 2b/3b stats:

Perry: 325 IP, 48 doubles, 11 triples
Marichal: 300 IP, 23 doubles, 5 triples

With such a spread between the 2 pitchers one can not argue "luck" was the difference.

OOTP replay avg of 5 sims of season:

Perry: 300+ IP, 45 doubles, 15 triples
Marichal: 300+ IP, 46 doubles, 17 triples

They seem like clones. Those extra base hits do make a difference in a pitchers stats. Pitchers do have some control over how hard they get hit, and not just the hits that go over the fence, but that is not part of the game engine for pitchers.

This is what I have meant in the past by pitchers being basically the same on a team. I see the same thing in my 2 fictional leagues.

So, while Perry actually looked close to what he really did in all 5 replays, Marichal never looked like Marichal.

The DIPS engine tends to make the worst pitchers look a lot better than they actually were since much of what made them so bad isn't part of the pitcher rating system, and makes a lot of the best pitchers look worse.

I don't plan on getting into a debate about DIPS. I'm just pointing out how the DIPS game engine has messed up historical pitcher stats. In the other game I play for historical, everything that every player (batters & pitchers) did is accounted for in the game engine and thus does a vastly better job of representing the individual players.

I will say again, however, that overall league stats, team stats, standings, and individual batter stats have improved dramatically. When I see the Mets actually win the NL East with 105 wins, that in itself was a remarkable improvement since previously they never finished better than 4th and never at or above .500. And seeing the Expos actually finish last in four sims and 5th in one sim instead of winning the WS or finishing in the top 3 in the NL East the rest of the time was great.

If the pitching stuff ever can get fixed I will probably go back to OOTP for historical because I like the game play and extras that OOTP offers that other games don't.

11-28-2010, 01:22 PM	#1
StyxNCa Hall Of Famer Join Date: Dec 2004 Location: Victoria, Texas Posts: 3,136	Historical Pitching Accuracy I simmed the 1969 season 5 times the other day and I was pleased with the overall accuracy of the game. The standings looked pretty good and the individual batter stats looked pretty good. However, the problem I had with the individual pitcher stats is still there for the most part. Many pitchers I looked at had stats that didn't come anywhere close to what they actually did. This has been the problem since the DIPS engine came into play. There is no way pitcher stats will ever look right with this engine. I am not intending to open an old argument here about DIPS, but saying pitchers don't control the number of doubles and triples they allow makes for some stats that are way off what they actually did. Just these two examples: 1969 SF pitchers, Gaylord Perry and Juan Marichal. Real life 2b/3b stats: Perry: 325 IP, 48 doubles, 11 triples Marichal: 300 IP, 23 doubles, 5 triples With such a spread between the 2 pitchers one can not argue "luck" was the difference. OOTP replay avg of 5 sims of season: Perry: 300+ IP, 45 doubles, 15 triples Marichal: 300+ IP, 46 doubles, 17 triples They seem like clones. Those extra base hits do make a difference in a pitchers stats. Pitchers do have some control over how hard they get hit, and not just the hits that go over the fence, but that is not part of the game engine for pitchers. This is what I have meant in the past by pitchers being basically the same on a team. I see the same thing in my 2 fictional leagues. So, while Perry actually looked close to what he really did in all 5 replays, Marichal never looked like Marichal. The DIPS engine tends to make the worst pitchers look a lot better than they actually were since much of what made them so bad isn't part of the pitcher rating system, and makes a lot of the best pitchers look worse. I don't plan on getting into a debate about DIPS. I'm just pointing out how the DIPS game engine has messed up historical pitcher stats. In the other game I play for historical, everything that every player (batters & pitchers) did is accounted for in the game engine and thus does a vastly better job of representing the individual players. I will say again, however, that overall league stats, team stats, standings, and individual batter stats have improved dramatically. When I see the Mets actually win the NL East with 105 wins, that in itself was a remarkable improvement since previously they never finished better than 4th and never at or above .500. And seeing the Expos actually finish last in four sims and 5th in one sim instead of winning the WS or finishing in the top 3 in the NL East the rest of the time was great. If the pitching stuff ever can get fixed I will probably go back to OOTP for historical because I like the game play and extras that OOTP offers that other games don't.