John Heyman made a comment during the MLB Network Hall of Fame show that cut both ways for me. Rich Lederer, long time advocate for Bert Blyleven's Hall of Fame induction transcribes Heyman's remarks:

"I never thought [Bert Blyleven] was a Hall of Famer when he was playing, and I saw him play his entire career."

"[His popularity] is based on a lot of younger people on the Internet who never saw him play."

"It's not about stats...it's about impact."

- Jon Heyman on MLB Network, 1/12/09

Patrick Sullivan, another writer at Baseball Analysts, looks into poor award voting and its effects on Hall of Fame ballotting. Bert didn't do well in the Cy Young voting, and that's held against him as well.

The question for me is, why is there such a huge difference between the stats and the perception? This query is especially important to me, because to a certain extent I sympathize with Heyman on this one. Unlike Heyman, I can be convinced otherwise.

The Minnesota factor doesn't ring true to me. Yes, it was a small market, but growing up during the 1970s I was well aware of players like Tony Oliva and Rod Carew, both who more than got their due winning batting titles. With 12 teams in the AL at the time, the Twins would play the big markets in the east four times during the season, giving someone like Blyleven the possibility of three or four starts. Sports writers in the east saw him pitch more than enough. They just weren't impressed with what they saw.

The biggest reason stems from the disconnect between Blyleven's ERA and his winning percentage. Were his stats deceiving? Was there something about the distribution of his runs allowed that is fooling the "younger people on the Internet who never saw him play?"

I decided to look at games scores in a probabilistic fashion. Given a particular game score, or range of game scores, what is the probability of winning that start? The probability of losing that start? If Blyleven's expected wins and losses are in line with his actual wins and losses, then the people unimpressed when they saw him play have a point. If his game scores indicate Bert should have won a lot more, or lost a lot less, then his Hall candidacy looks much better. I took all games in the Day by Day Database, currently from 1957 through 2008, and built the following model, using Game Score divided by 10 to smooth things out.

The two percentages are per start, not per decision. Now this is not a perfect model. It should probably be adjusted for era, but with over 50 years of data, I suspect those things even out. With this data, we can take each individual start, add up the probabilities for each game score, and get expected wins and expected losses. I did this for every pitcher who collected 200 starts between 1957 and 2008. I also calculated the ratio 100.0*Wins/PredWins and 100.0*Losses/PredLosses. A value over 100 for wins indicates the pitcher won more game than expected. A value over 100 for losses indicates the pitcher lost more game than expected.

Note that at very high game scores, pitchers don't win that often. That seems counter intuitive until you realize that really high game scores often happen in extra-inning games. Once the scores get into the 80s, there's a nice downward progression of winning percentage, and a nice upward progression of losing percentage. A score of 50 is neutral, and you can see that above 50, pitchers win more often than they lose, and below 50 pitchers lose more often than they win.

Three hundred fifty five pitchers qualified, and here they are the top 20 ranked by Win Ratio. For the full list, click here.

Blyleven ranked 309th in Win Ratio, 92.46. He started 685 games, and compiled a record of 286-248 in those games. His expected won-loss record was 309-205. The loss number is somewhat more interesting. Not only did Blyleven fall short on wins, but his ability to go deep in games while his team didn't score saddled him with an extra 43 losses, a Loss Ratio of 121.03. If he puts up those 286 wins with 220 losses, my guess is that he would have been in the Hall a long time ago.

The person just ahead of Bert in Win Ratio is Don Sutton. Don compiled a 321--253 record in his starts, when he should have been 347-224. By reaching the magic 300 level, Don made the Hall, but there were quite a few people who felt the same way about Sutton that they did about Bert.

Of course, next to Nolan Ryan, Bert can't complain too much. Ryan ranked 348 out of 355, and his predicted record was 384-209! (He compiled a 318-291 record in his starts.) That's a Win Ratio of 82.91 and a Loss Ratio of 138.97. The only person in the study with a worse Loss Ratio than Ryan was Bob Gibson at 139.71. My guess is if we adjust for era and parks, Gibson's prediction might be more in line with his record.

Of the 300 game winners in the study, only one pitcher is shown as being undeserving, Tom Glavine. His record of 305-203 in starts should have been closer to 267-232. Glavine will go down as the anti-Blyleven, someone good but not great who makes the Hall because he played on very good teams during his career.

In conclusion, it's fairly obvious that Blyleven pitched well enough to win 300 games, and other factors not only kept his win total low, but greatly inflated his loss total. Seen in this light, there's no doubt Bert should get the Hall call.

Update: See an interesting graph based on the data here.