Secret Formula 2006

Postby Mean Dean » Wed Jan 31, 2007 11:25 am

[quote:727a47b2d0]HAL's triggers/thresholds to execute certain strategies are quite safe IMHO. Derrek Lee won't Hit ad Run if you leave him alone. No matter if it's Figgins and Podsednik and Freel all rolled into one with a SS rated 4 defensively, poor pitcher HLD and weak catcher T. Even if you tag him to H&R more and set H&R to aggresive HAL won't do it. So my logic behind those "liberal" settings is that if I know HAL won't goof up leaving it alone and if on top of that I find an underperforming pattern when I tinker with settings heavily, why mess with it? [/quote:727a47b2d0]
[b:727a47b2d0]Cummings[/b:727a47b2d0], I think this paragraph makes a lot of sense and is a good point. The question is what conclusions it necessarily leads to. My questions for you -- I don't know the answers, but they're what I think we need to establish -- are:[list:727a47b2d0][*:727a47b2d0]Okay, so it's not going to H&R with Lee (great hitter, "D" H&R) even on an aggressive setting. I buy that... but who [i:727a47b2d0]is[/i:727a47b2d0] it H&R'ing with, then? Is it only H&R'ing with the Jack Wilsons, poor hitters with "B" H&R? Or will it also do it with the Tad Iguchis (decent hitters).... the Derek Jeters (excellent hitters, but with little power)... the Alex Rodriguezes (great hitter with power)... all those guys are "B"s, so, where is it drawing the line? And will it ever H&R with a "C" or "D"?[*:727a47b2d0]The same questions as #1, relating to bunts.[*:727a47b2d0]Isn't it an altogether different story with steals? I can easily imagine HAL stealing in iffy situations (and in fact, people complain pretty regularly here about him doing so.) What kind of SB totals and percentages do you get?[/list:u:727a47b2d0]
Mean Dean
 
Posts: 55
Joined: Tue Jul 03, 2012 2:34 pm

Postby Play By The Rules » Wed Jan 31, 2007 12:48 pm

My opinion is that 90% of H&R has to do with the #2 hitter. (Bunting too, in addition to the pitcher.) Since the #2 hitter by principle shouldn't be weak, my default setting is "conservative" at best, and I move it to "normal" only when facing a team with 1s up the middle in a situation where I feel getting on the bad side of the pitchers' card (vs. your Redmans and Lidles but even moreso your Padillas) is advantegeous.

I also closely track my teams' H&R batting average and if it is above the card of the #2 player in question I am more inclined to use H&R.
Play By The Rules
 
Posts: 55
Joined: Tue Jul 03, 2012 2:34 pm

Postby cummings2 » Sun Feb 04, 2007 3:27 pm

[b:be64a073f5]J-P,[/b:be64a073f5]

Sorry to see a good team struggle so much, between injuries and underperforming it's tough. I have had some teams like that, shucks big toes but...still we keep hoping it'll get better.


[b:be64a073f5]Dean,[/b:be64a073f5]

Sorry for the late reply to your post, I had been checking in page 4 of this thread to get to J-Pav's team and had stopped monitoring the front. OK, good questions, exactly what I've been going after. Where I stand at the mo (I'm almost sure I've written this before) is that HAL's decision making process is linear as opposed to comparative. Two reasons why it's important to make this distinction:

1. If its linear and not comparative it is very likely that HAL will [i:be64a073f5]not[/i:be64a073f5] choose the best strategy in relation to all the other options but that the strategy that s used will have some sort of success rate that HAL deems acceptable [i:be64a073f5]given the settings the team is running under[/i:be64a073f5]

2. If it is linear then the sequence of that linear "thinking" is quite imprtant to figure out.

My impression is that the thinking goes:

Is there a man OB?
if Y= Do I steal (Y/N) -squeeze, delayed, double steal, etc
if N= Do I Bunt (Y/N)
if N= Do I H&R (Y/N)
if N= Swing away

I really don't want to hijack the thread with all of this which I guess I should post in the strategy section but, to answer you Qs:

1. In my experience HAL [u:be64a073f5][i:be64a073f5]will[/i:be64a073f5][/u:be64a073f5] H&R with "C" hitters -especially once they've been upgraded to B and the runner being held on 1B has good chances of getting the lead but lower chances to steal safely. For example Eddie Garabito's *4-7/11 (15-10).

2. Who HAL opts to H&R with? It seems to be quite related to SLG. High BA low SLG (like Lofton) seem to be the most affected. Of the examples you mentioned unfortunately Jeter and Iguchi are the most likely to be affected. Unlikely that Wilson will have good runners on base (given his usual low batting spot) and I really doubt that A-Rod would H&R. I have some reason to believe, though not enough data to back this up, that for whaever reason high GbA hitters do seem to H&R more, so of the ones you listed, who is HAL most likely to H&R with? I'd say 1. Iguchi 2. Jeter 3. Wilson 4. A-Rod. [[b:be64a073f5]Note[/b:be64a073f5]: [i:be64a073f5]I [/i:be64a073f5] personally think that giving HAL the benefit of the doubt that it does take into consieration GbAs is giving HAL too much credit but as I said there is some consistant data to back up at least looking into that idea]

3. You are right, steals are quite different, IMO HAL seems to be quite swayed by the "getting lead safely chances" and the "safe stealing with lead chances" it is farily easy to deduce that HAL does take into consideration the pitcher's hld and catcher's arm. Manager settings seem to give HAL the green light if the thresholds are met that it deems to be "conservative/agressive, etc" Normal seems to be around 70-72% Safe chances (With results/rolls variating from to 66-68% and up to 76-78%)

4. Bunting, I haven't really followed bunting as closely as stealing and H&R but my impression is that bunting succ. rate for HAL is around 90% for normal. and depending on the situation (0 outs, tied game) HAL will bunt no matter who it's with. Brian Roberts seems to bunt much more than it should -but as I said before I haven't followed Bnt/Sqz.

Once again, I apologize for the late reply, and for hijacking the thread momentarily. I'll post some of the raw data in the Strategy section or just e-mail it to whomever is interested maybe someone else will see something I am missing.

-Hope it made some sense.

C2
cummings2
 
Posts: 55
Joined: Tue Jul 03, 2012 2:34 pm

Postby cummings2 » Sun Feb 04, 2007 4:51 pm

quick note about what I just wrote:

[quote:31e26ad928]HAL will bunt no matter who it's with[/quote:31e26ad928]

*** If the hitter is rated A for bunting.
cummings2
 
Posts: 55
Joined: Tue Jul 03, 2012 2:34 pm

Postby Mean Dean » Sun Feb 04, 2007 6:02 pm

[b:781ee2a7b9]Cummings[/b:781ee2a7b9], if your understanding is the case -- that it will H&R more with the high-gbA hitters, and won't bother to H&R with the great hitters -- then that does suggest an aggressive H&R stance, since that would basically amount to saying that HAL uses the H&R intelligently. The only thing you'd have to do is set guys like Jeter, who are good hitters despite low SLG, to don't H&R. The problem, of course, is establishing that it [i:781ee2a7b9]does[/i:781ee2a7b9] work like that, when we can't really tell when the game is H&Ring. PC game tests might help here; I assume it must have that data, although actually I can't remember seeing it, offhand...

I am not too nuts about a 70-72% safe chance on steals. I get very high percentages by putting the setting on conservative (I do think extra conservative is too risk-averse), and the individual stealers on steal more. This in all probability does result in fewer attempts, but depending on who exactly the basestealers are, it still can be a very substantial number of attempts. If it's a choice between 100 attempts at an 85% chance or 200 attempts at a 70% chance, I'd have to figure the first one is better. (If we count a CS as twice as bad as a steal, 85 SB/15 CS would be 55 "net steals", and 140 SB/60 CS would be 20 "net steals.") So I am still not sold that aggressive SB is a good tradeoff, unless the extra SB attempts really avoid a ton of DPs, or my guesses about the effects of the settings are off.
Mean Dean
 
Posts: 55
Joined: Tue Jul 03, 2012 2:34 pm

Postby J-Pav » Fri Feb 09, 2007 10:33 pm

With the Tour Championship finally underway, I browsed over the lineups to see if [i:101feabfbb]The Secret Formula[/i:101feabfbb] could be found among the tour finalists.

In what has to be the ultimate irony, [b:101feabfbb]luckyman[/b:101feabfbb] appears to be the closest with just over $34 mil spent on pitching (well suited to his park and with the four highest priced pitchers seeing the four highest numbers of innings pitched), 1-1-2 at 2B-SS-CF, and a salary construction of 11, 8.5, 4.5, 4, 3.5 ($32 mil total) for his top five offensive players.

Next would be [b:101feabfbb]mcsoupy[/b:101feabfbb] with just over $32 spent on pitching, 2-2-2 up the middle, and $7, 6, 5, 5, 4.5 for the top five hitters (somewhat low at $29 mil total, but well spread across the rest of the team). This set up is nothing new to the [b:101feabfbb]soup[/b:101feabfbb], as his were the tour teams I used in my earlier example (see page two, 22 Dec 2006 15:45).

[b:101feabfbb]Cminton[/b:101feabfbb] and [b:101feabfbb]qksilver[/b:101feabfbb] are also there; however, they're a little uneven on their pitching salary construction with [b:101feabfbb]qk[/b:101feabfbb] going for budget number three and four starters and [b:101feabfbb]Charlie[/b:101feabfbb] trying five or six [i:101feabfbb]all[/i:101feabfbb] "strong versus lefties" SPs with almost no bullpen to speak of.

In the East, [b:101feabfbb]Odieworks[/b:101feabfbb] went just under $32 mil on pitching (spread out well SP/RP and suited to his park), [color=red:101feabfbb]3[/color:101feabfbb]-2-2 up the middle and $9, 7, 6.5, 6, 6 (for $35 mil). A one or a two at 2B would've made this team the best Secret Formula example. The [b:101feabfbb]LMBombers[/b:101feabfbb] are not far off, but spent four mil less than Odie on the top five offensive players and are 3-2-2.5 (with the CF platoon) up the middle.

As I watched the draft unfold, I forecasted that the West would be by far the strongest division. At first glance, I thought both [b:101feabfbb]Kaviksdad[/b:101feabfbb] and [b:101feabfbb]Spicki[/b:101feabfbb] had very strong teams; but, now I'm wondering how it will look at the end of the season when they both spent the least ($28 mil and $26 mil) on their top five hitters and went very heavy on pitching ($36 mil and $38 mil).

In the food for thought area, the teams that spent the most on their top five offensive players [b:101feabfbb]C2[/b:101feabfbb] ($45 mil), [b:101feabfbb]Odieworks[/b:101feabfbb] ($35 mil) and [b:101feabfbb]Sknsfan[/b:101feabfbb] ($35 mil) all find themselves in first place in the very early going, with [b:101feabfbb]qksilver[/b:101feabfbb] (also $35 mil) waiting for his wave. I started this year's thread because I was very intrigued by my salary construction ideas. Even though the thread spent way too much time on pitching salaries (i.e., the number $32 mil), I'll be excited to watch how these salary balances will play out, especially [b:101feabfbb]C2[/b:101feabfbb]'s $21 mil spent on pitching, as the season evolves.

FWIW, these twelve teams have 51 rings between them (led by [b:101feabfbb]C2[/b:101feabfbb]'s [i:101feabfbb]nine[/i:101feabfbb]), so my observations should be taken with a grain of salt as any one of these guys could be hiding [i:101feabfbb]their[/i:101feabfbb] secret formula for the Final Battle. It will be fun watching it all unfold...

:o
J-Pav
 
Posts: 55
Joined: Tue Jul 03, 2012 2:34 pm

Postby J-Pav » Sat Feb 10, 2007 12:15 am

My tour season is over, but it's five below zero here in the Chicagoland area, so I can't turn to my golf game yet. That and my intense desire to raise my community points by 0.04 more (sarcasm here) brings me back to the keyboard.

Thinking some more about salary construction, I came up with this idea. The average first nine offensive salaries in order from highest to lowest (for the Tour Championship) are 9.33, 8.07, 6.10, 4.96, 4.46, 3.93, 3.07, 2.42 and 1.76.

I don't know how predictive this might turn out to be (there are definitely some drawbacks to this notion), but going down the line I placed a minus sign (-) if that team was below the average at that rank and a plus sign (+) if that team was above the average, for the highest nine hitters' salaries on each team in descending order. Many times a team was just a touch above or below the average (this is the drawback part), but rather than create a third sign I just went plus or minus for a quick visual.

odieworks - - + + + + - - -
stevep107 - - - - - - + + +
maligned + + + - - - - + +
LMBombers - - - - + + - + +
Cummings2 + + + + + + - - -
mcsoupy - - - - + + + + +
qksilver - + + + + - - + +
cminton - - - + + + + + -
sknsfan + + - - - - - - -
marcus wilby + + - - - - - - -
kaviksdad - - - - - + + - -
spicki17 - - - - - - - - +

As you can see, [b:7d3c73a092]Cummings2[/b:7d3c73a092] and [b:7d3c73a092]qksilver[/b:7d3c73a092] each have six positives, [b:7d3c73a092]maligned[/b:7d3c73a092], [b:7d3c73a092]mcsoupy[/b:7d3c73a092] and [b:7d3c73a092]cminton[/b:7d3c73a092] each have five. (Note: [b:7d3c73a092]mcsoupy[/b:7d3c73a092] was actually very close to having seven).

I'm wondering if it would be a worthwhile exercise to strive to be just above the average in all nine categories, and not just the first five. It's obviously not practical to fore go a certain player (who may be a bargain) just because he's 0.02 below the average in a category. But [i:7d3c73a092]in theory[/i:7d3c73a092], if you're just a bit stronger than the average manager in every lineup position across the board, doesn't it make sense that you're more likely to win more games?

Although the pitching salaries are equally if not more important, I'll spare everyone going through them, as hopefully the point has been made. Any thoughts?
J-Pav
 
Posts: 55
Joined: Tue Jul 03, 2012 2:34 pm

Postby J-Pav » Sat Feb 10, 2007 1:04 am

Alright, I can't stop myself now, so here are the pitchers. I took the first seven salaries, which averaged 7.39, 5.56, 4.55, 3.82, 2.91, 2.48 and 1.81.

odieworks - - + + + - +
stevep107 - - - + + + +
maligned + - - - - - +
LMBombers - - - - + + +
Cummings2 - - - - - - -
mcsoupy - - + + + + -
qksilver + + - - - - +
cminton + + - - + + +
sknsfan - - - - + + +
marcus wilby + + + + - - -
kaviksdad + + + + + + -
spicki17 + + + + - + +

Having laid out the visual, let it be said that there are an awful lot of intangibles that aren't reflected in adding 12 numbers and dividing by 12. The negative signs just below the average and the positive signs just above the average aren't factored in; star (*) SPs, SPs and RPs aren't distinguished (i.e., so there may be dollars spent inefficiently on pitchers not getting the innings to justify their salary); "bargains" aren't accounted for in any way; and lastly, maybe only the first five salaries matter much (or six, or eight, or nine, etc), not the top seven.

Still, it's something more to consider when you're trying to build your team.
J-Pav
 
Posts: 55
Joined: Tue Jul 03, 2012 2:34 pm

Postby cummings2 » Sat Feb 10, 2007 1:17 am

[quote:2168c4b1ea]Cummings2 - - - - - - - [/quote:2168c4b1ea]

...My makeshift array of flatliners :lol:

Kidding aside, I see a lot of merit into this way of seeing a team. As you have pointed out there are some issues to consider but on the whole it's certainly worth looking into it.

Very cool J!

-P.S. Sorry the suggestions didn't really help your team posted earlier. Tough stuff there.
cummings2
 
Posts: 55
Joined: Tue Jul 03, 2012 2:34 pm

Postby cummings2 » Sat Feb 10, 2007 1:33 am

[b:4234da27de]Dean[/b:4234da27de],

Your rationale in re: to SB attempts/succ. rate makes much more sense than my convoluted thinking. especially this:

[color=blue:4234da27de][i:4234da27de] it's a choice between 100 attempts at an 85% chance or 200 attempts at a 70% chance, I'd have to figure the first one is better. (If we count a CS as twice as bad as a steal, 85 SB/15 CS would be 55 "net steals", and 140 SB/60 CS would be 20 "net steals.")[/i:4234da27de][/color:4234da27de]

I'll gather some of the info I have for SBs and H&R and start a thread over in the strategy forum, I'd be very interested in seeing the numbers from a different perspective -with your cooperation. Thanks for pointers :D
cummings2
 
Posts: 55
Joined: Tue Jul 03, 2012 2:34 pm

PreviousNext

Return to Strat-O-Matic Baseball Online 20xx

Who is online

Users browsing this forum: No registered users and 2 guests

cron