Top team in my league only has 4 relievers?

[Sashindor]

Is this some weird coincidence or is he/she onto some hot new strat I don't understand? Granted, I am in a Bronze League, ha, but his/her team is currently 97-45

[<Pion>]

Did you look at their roster page and see that, or from the team page where it displays both hitters and pitchers on the right side. If the latter, I know that when I experimented bullpen by committee and made everyone a middle reliever, it would only show 4 relievers on the team page even though I had six designated as middle relievers. But it could be that they went with more hitters or just did not assign roles for some.

[daves]

I have seen teams with nothing but relievers and just one starting pitcher. The starting pitcher had a low stamina and was a lefty, with a bunch of righties in the pen..
There are crazy strategies out there.

[Charlatan]

Originally Posted by daves

I have seen teams with nothing but relievers and just one starting pitcher. The starting pitcher had a low stamina and was a lefty, with a bunch of righties in the pen..
There are crazy strategies out there.

That sounds like a opener-follower strategy.

[herbyhancok]

In iron last week and bronze this week I have had teams in my league that only run 5-6 relievers(with standard five-man rotations) that are constantly tired, yet have one of the best bullpens in the league. I check their rosters occasionally, they're definitely not rotating relievers between the active and reserve rosters. I never played PT 1, but at this point I'm curious now whether it has ever been confirmed whether tiredness even has an effect in PT?

[Kushiel]

Originally Posted by herbyhancok

In iron last week and bronze this week I have had teams in my league that only run 5-6 relievers(with standard five-man rotations) that are constantly tired, yet have one of the best bullpens in the league. I check their rosters occasionally, they're definitely not rotating relievers between the active and reserve rosters. I never played PT 1, but at this point I'm curious now whether it has ever been confirmed whether tiredness even has an effect in PT?

I use an 11 man staff and have since the Beta last year. How you build your bullpen will control if they are over worked. Most players that have been playing since November use one starter or more in the bullpen. Personally I prefer 4 starters and 2 closers in my bullpen. I have little to no trouble having a rested bullpen and have never been slaughtered like the scores shown in another post.

Edit: I meant to add that that the 12th pitcher spot that many people think they need is much better used in your platoons as you rise in levels.

[atabakin]

In PT 1.0 I used (still use, I suppose) a 5-man bullpen. Pitcher fatigue has been cranked up in this year's version, so even with a stable of starters in the bullpen that can't work now since pitchers (both starters and relievers) recover quite a bit slower.

[herbyhancok]

Thanks for the responses, but it still doesn't address my question. Has anyone been able to reasonably suss out whether pitcher fatigue even effects performance in PT? The teams I've seen do this haven't be using SPs in relief, it's just 5-6 relievers. There's no clever usage going on, just 5-6 guys that are completely gassed(30% or below) at all times, except for after days of rest and they're leading the league in bullpen ERA.

[Charlatan]

My bullpen is like that but it's more of a result of having 5 starters that are ok (but not great) and having 7 relievers all with stamina between 14-17. Which means the AI only rarely uses them for more than one inning. Which means my guys are always pitching. So they're always tired. Now they can catch up on the occasional quality start or a day off, but when crunch time comes they get awfully tired.

I do not know for sure if it affects their performance. I go on the assumption that it does.

[allenciox]

So I recently did a regression analysis predicting ERA+ for pitchers for the first few seasons for three different leagues in PT 20. I included as potential predictors all the different pitcher ratings, and performed the analysis separately for starters and relievers. For starters, only 3 ratings played any part in predicting their ERA+, and they are the ones you would expect, in order of importance: CON under 50, MOV , STU, CON over 50. (The model found that additional points in CON over 50 were worth only half as much as points below 50).

For relievers, the same results held up with one exception: STM (stamina) was an extra variable in the model, significantly less important than the other three but very statistically significant (p < .0001). Relievers with higher stamina ratings did better holding all other attributes constant. That suggests to me that the game does penalize tired pitchers (and also suggests that it generally does consistently substitute for SPs when they get tired).

[herbyhancok]

Thanks for the info. I'm not the most knowledgeable, it's been years since I've taken a stats course and most of what I know's eroded away. That said, wouldn't a p-value that small be too insignificant to draw any conclusions from?

[Dogberry99]

Originally Posted by herbyhancok

Thanks for the info. I'm not the most knowledgeable, it's been years since I've taken a stats course and most of what I know's eroded away. That said, wouldn't a p-value that small be too insignificant to draw any conclusions from?

Actually, as I understand it, this - what constitutes an appropriate level at which significance can be determined - is one of the hallmark differences between statistical analysis in the so-called "hard" sciences (ie Chemistry, Biology) and "soft" sciences (ie Psychology, Sociology).

[allenciox]

A p-value of less than .0001 indicates how likely a relationship could be found by chance. It means that if I there really "wasn't a relationship" then there would be less than a 1 in a ten thousand chance that it would have been detected.

Now the software I am using only reports significance scores down to that level. In actuality it is probably much, much lower than that, which means that you can be virtually sure that there is a relationship between the reliever's stamina and their era.

[herbyhancok]

Originally Posted by Dogberry99

Actually, as I understand it, this - what constitutes an appropriate level at which significance can be determined - is one of the hallmark differences between statistical analysis in the so-called "hard" sciences (ie Chemistry, Biology) and "soft" sciences (ie Psychology, Sociology).

That definitely explains my misconception, I took business stats.

[Sashindor]

Originally Posted by allenciox

So I recently did a regression analysis predicting ERA+ for pitchers for the first few seasons for three different leagues in PT 20. I included as potential predictors all the different pitcher ratings, and performed the analysis separately for starters and relievers. For starters, only 3 ratings played any part in predicting their ERA+, and they are the ones you would expect, in order of importance: CON under 50, MOV , STU, CON over 50. (The model found that additional points in CON over 50 were worth only half as much as points below 50).

For relievers, the same results held up with one exception: STM (stamina) was an extra variable in the model, significantly less important than the other three but very statistically significant (p < .0001). Relievers with higher stamina ratings did better holding all other attributes constant. That suggests to me that the game does penalize tired pitchers (and also suggests that it generally does consistently substitute for SPs when they get tired).

Right on! I thought my thread was dead and buried but as usual, this community turns out to be super rad.