What will you go for in the playoffs?

I had a choice between my top two starters for the WS, they were both rested and available for key games.

Starter A
'71 Vida Blue, actual stats  OAV# .197, WHIP# 0.98, ERC# 1.92, BB/9# 2.49, HR/9+ 141
regular season stats  W-L 30-7, ERA 3.03, OAV.224, WHIP 1.14, 19HR in 360 IP

Starter B
'80 Mike Norris, actual stats  OAV# .205, WHIP# 1.04, ERC# 2.20, BB/9# 2.65, HR/9+ 143
regular season stats  W-L 27-7, ERA 2.24, OAV.207, WHIP 1.07, 13HR in 338 IP

So according to these numbers Blue had better actual stats but Norris had the better season. 
According to logic I still went for Blue to start the WS and he got beat up pretty bad to get the loss.
Norris came in game two and did very well to get the win.
Then in game 5 with the series tied-up at two I had both pitchers rested and decided to go with actual stats logic despite season and WS results, thinking the odds of Blue getting blown out twice were against actual stats logic... well, I paid for that choice as he got blown out even worst then the first game...

That's not the first time this happens, and I always hesitate to go with the guy who's had the better season, still thinking that no matter what happened in the season, the actual stats are what matters with the sim engine...

Any thoughts on that?
                          

wow good question, I'm a beginner in this game but I can try to put my two cents in. maybe Norris matches up better vs the team your playing, are they both right handers?  Norris era is lower, to me they are pretty similar. I wouldnt go back on your decision, I think you made a good one based on a gut feeling of which pitcher is better with actual stats.  I probably would have dont the same thing.

Posted by chisox378 on 8/13/2010 10:15:00 AM (view original):
wow good question, I'm a beginner in this game but I can try to put my two cents in. maybe Norris matches up better vs the team your playing, are they both right handers?  Norris era is lower, to me they are pretty similar. I wouldnt go back on your decision, I think you made a good one based on a gut feeling of which pitcher is better with actual stats.  I probably would have dont the same thing.

True that they are similar but Blue is a lefty and strikes out more batters and does have a slight edge in every categories of actual stats. But my opponents line-up had six switch hitters so I didn't think the lefty/righty option would really make a difference. And as for Norris matching up better, it could be but I just don't see that in stats...

Normalization answers a lot of the difference. AL teams in 1971 averaged 3.87 runs per game, 2nd-lowest from 1969 on; 1980 average was 4.51, third-highest from 1969-1984 Blue's stats all normalize upward, while Norris' all normalize down. I'd still expect Blue to have better seasons on average, but not by very much, and that would depend on cap. In higher caps, where offenses are loaded, Norris might be expected to perform better.

HR/game were 15% higher in 1980 than in 1971, which may explain why Blue's RL HR/9 of 0.55 translates to .72 in performance history and Norris' .57 RL results in .64 in perf. history. In a high-offense theme league, that would make a huge difference. ERC is not used by the sim engine, but as a summary of pitching effectiveness it's usefl. Blue's RL ERC of 1.81 normalizes (ERC#) to 1.92, an increase of 6%; Norris' 2.26 ERC normalizes to 2.20, a decrease of 2.6%.

Posted by joshkvt on 8/13/2010 12:16:00 PM (view original):
Normalization answers a lot of the difference. AL teams in 1971 averaged 3.87 runs per game, 2nd-lowest from 1969 on; 1980 average was 4.51, third-highest from 1969-1984 Blue's stats all normalize upward, while Norris' all normalize down. I'd still expect Blue to have better seasons on average, but not by very much, and that would depend on cap. In higher caps, where offenses are loaded, Norris might be expected to perform better.

HR/game were 15% higher in 1980 than in 1971, which may explain why Blue's RL HR/9 of 0.55 translates to .72 in performance history and Norris' .57 RL results in .64 in perf. history. In a high-offense theme league, that would make a huge difference. ERC is not used by the sim engine, but as a summary of pitching effectiveness it's usefl. Blue's RL ERC of 1.81 normalizes (ERC#) to 1.92, an increase of 6%; Norris' 2.26 ERC normalizes to 2.20, a decrease of 2.6%.

I see what you mean, and I usually try to draft players who normalize downard but I always thought that in the end the # stats was what mattered, if a pitchers' WHIP# was 1.00 regardless if it was up from a 0.96 or down from 1.04 WHIP it would still give you a 1.00 WHIP#...

I'm finding out that some of the teams that use the base-stealing switch hitters ala Raines, Wilson, etc,...those players mentioned crush left handed pitching and have a harder time vs. righties. 

Chisox, theres a quality difference between LH and RH pitchers. A hitter will have to face Joss and guys of that caliber a lot during the season. Lefties of that caliber don't exhist. When I draft a lefty, its usually a long reliever that just happens to be LHed. Definitly nobody I'd count on. So those switch hitter splits can be skewed.

The # stats are useful search tools, but I don't think the sim engine uses them for anything. It appears that the sim puts a stronger emphasis on adjusting stats based on the player's season/league than does whatever formula is used to create the # stats. 

In general, I think you also need to look at ballpark/opposing team.  If your guy pitched well/crappy at home/on the road, then there may be a reason for that pattern.  As well, if one gave up a lot of bombs, for example, and you're facing a HR heavy lineup, you may want to go with the other guy.