Where is that thread about bad pitching decisions- look here

[Semper Gumby]

Okay, I'm getting beyond frustrated with the game engine logic for pitching -

HAL has no qualms about yanking Pedro early but he'll keep Cliff Lee in the game -

[code:1:f9ead1d33f]United IP H R ER BB SO HR PC ERA
B.Colon WIN(9-5) 6 12 1 1 0 4 1 107 2.66
H.Street HOLD(5th) 2 1 0 0 0 0 0 32 3.46
F.Rodriguez 1 0 0 0 0 0 0 9 4.09
Totals 9 13 1 1 0 4 1

Jarheads IP H R ER BB SO HR PC ERA
C.Lee LOSS(1-3) 9 10 5 5 0 6 3 147 4.71
Totals 9 10 5 5 0 6 3[/code:1:f9ead1d33f]


Lee's individual settings are IBB Less / Quick Hook.


And, please also review the pitching logic in the 80s engine too - here is the most recent version of "Why HAL?"

[code:1:f9ead1d33f]Devil Dogs IP H R ER BB SO HR PC ERA
J.Reuss 7 7 3 3 0 6 1 105 2.05
D.Bair LOSS(3-3) 1 2 1 1 3 1 0 37 3.69
W.Hernandez 0 1 0 0 0 0 0 5 1.50
Totals 8 10 4 4 3 7 1[/code:1:f9ead1d33f]


Reuss' settings are SLOW HOOK / IBB Less and you can see from his ERA after 45 games - he has a decent card at work here.

I've grown from amused to the end of my SOM rope. I've documented more and more of these instances (80s version is harder since we rarely know the card). It appears HAL loves to play havoc with the pitching and merely enters the game without any consistent patterns.

If you can't program the pitching logic using the manager / individual settings, why bother offer them. :x

[MARCPELLETIER]

Semper,

I don't know the code, but I do think that Hal is "roughly" consistent, and your data show that consistency.

As a rule of thumb, when a game is tight (equal or less than one run difference) Hal is likely to change the SP for a reliever if
-a runner gets on-base
-a reliever offers a better face-to-face scenario against the upcoming hitter than the SP
-relief can go deep (lots of fresh relievers available) OR you're getting close to the pitcher's number of innings before tiredness

With this rule of thumb, I can explain many of the moves Hal did with your team.

(Given you've changed your bullpen roster a few times during the season, I can't go into exact details)

Game 40: game is tight, getting close to S7, a runner gets on-base, Riske is better vs rhh than Prior (Atkins)
Game 60: game is tight, bullpen is deep, a runner gets on-base, Mullholand is better vs lhh than Haren (Sizemore, Counsell)
Game 49: game is tight, bullpen is deep, a runner gets on-base, Riske is a better match-up vs Cirillo than Pedro (Cirillo is weak, so Riske's HR not at risk; and Riske has less doubles and triples than Pedro, so less chances of allowing a run)
Game 45: harder to justify given that no one went on-base. But Pedro is in his S7 and has allowed two runners in the sixth, so he was close to his tired point. And just like game 45, you were losing by one run, and a weak right-handed hitter was coming at-bat. So a lot of consistency with game 45. Perhaps Hal preferred Politte---and a chance of walk---to Pedro's chances of doubles and triples.

Please note that in ALL preceding games, you were losing by exactly one run when Hal made his move. Pretty consistant.

As for the other games:
Game 25: Game is tight, one runner is allowed, you're leading by two, it's the 8th inning, in one word: it's a set-up/closer job. Plus, Wagner offers a much better face-to-face scenario vs Schneider than Pedro (which was nullyfied by the pinch-hitter Cameron)

Game 53: You were leading by two runs most of the way. Pedro allowed almost no men on base. Hence very few occasions to remove him. Relief was in danger to be thin with all bullpen participating in game 52. But Wagner came because he has the closer job (probably would have not come if he wasn't set as closer). Explains why Pedro stood all along until the 9th.

Game 9: same environment as game 53. Same result. Explains why Pedro stood all along until the 9th.


GAme 13: game has been tied all the way. Pedro has never been one hitter away of becoming tired. No right-handed weak hitter, so no risk of having Riske as a better match-up than Pedro. Explains why Pedro stood all along.


Game 33: not sure what is your complaint, but Pedro was one out away of his maximum. One runner got on-base. Cabrera is weak vs lhp, but N vs rhp, so Wagner potentially a better match-up. Later, Wagner was getting in his third inning, Politte was a better match-up vs Crosby.

The last game with Lee. This is a case where you were losing by three runs or more early in the game---putting Hal in a conservative mode, given your settings-----AND Lee has never been close of becoming tired (no hit/walk in both 7th and 8th, only one hit allowed in the 9th). So Hal never had to activate the "quick hook" mode.

In my eyes, there is solid consistency:
--Hal triggers the run with more rapidity when losing by one run. When losing by one run, particular with a man getting on-base, Hal definitively looks for the best potential match-up, which could trigger an quicker hook than usual. As if Hal was doing his best to not let the game go off hand. Call this over-managing, but it's pretty consistent.

(in my experience, this is true mostly with conservative settings. With aggressive settings, I feel that Hal is aggressive down by one run or leading by one run)

-- Conservative mode is perfectly illustrated by the fact that Hal leaves the starter there when losing early or when winning by two runs or more, particularly if no one gets on-base, until we're getting close to the tiredness point.

[Semper Gumby]

Marcus,

Unfortunately, I have quite a few examples of HAL's logic at odds with your observations especially when he allows pitchers to toss more than 7 innings.

If the manager and individual settings are to mean anything, you should expect that HAL's urge will be overridden.

Otherwise, Pedro is, by far, the best pitcher in this league and would have a few more wins had HAL allowed him to stay in.

Besides - changing Pedro out has little to no impact on whether my team scores any runs.

On the other hand, had I employed aggressive settings and normal hook for my starters I'd expect 7 -8 innings per start. In my current set-up, I am getting 7-8 innings per start.

Oddly, HAL even pulls WAGNER after one or two batters.

My observation - HAL's logic isn't based on the CD-ROM's advance program so we can't be sure what HAL will do on any game's events.

[Semper Gumby]

I'm beyond amused now so explain this one to me.

Why does HAL pull Pedro but allows Haren to toss away?

Same series

Pedro - SLOW HOOK / IBB Less / Conservative Relief


[code:1:d7c7aaefbb]Jarheads IP H R ER BB SO HR PC ERA
P.Martinez 7 2/3 3 0 0 0 10 0 91 1.45
B.Wagner WIN(2-2) 1 1/3 0 0 0 1 2 0 21 1.64[/code:1:d7c7aaefbb]


Haren - IBB Less only / Conservative Relief

[code:1:d7c7aaefbb]Jarheads IP H R ER BB SO HR PC ERA
D.Haren LOSS(4-7) 9 5 4 4 0 7 1 122 4.18
Totals 9 5 4 4 0 7 1[/code:1:d7c7aaefbb]

[DAVIDGAMBLE]

Did those three hits off of Pedro come in the same inning after he reached his POW? If they were spread out, you have a valid point, but if they came in the same inning, especially late, even on Slow Hook and conservative settings, Hal has to at least consider putting in a reliever if the game was tied or close.

As far as Haren, there's still no game context here...his hits and runs were spread out, he didn't appear to be shalacked and since you don't have QUICK hook on him (just the absence of SLOW hook) then unless he started getting raked after his POW inning there's no impetus to use a reliever here.

Once again though, I think if Pedro's three hits were spread then it was certainly not a great idea to relieve Pedro at that time given your settings, and you'd be very right. However if they WERE all in the same inning #1) give Pedro Kudos for throwing a no-no/perfecto through seven and #2) slap your hitters around for not having a big enough lead that Pedro needed to be yanked in that late situation

SwisherSweets