Page 4 of 6

PostPosted: Thu Feb 23, 2006 4:43 pm
by cummings2
O.K. after a slight H&R break, I have my second H&R team gettng ready to go.

I am using this next team to try a few slight changes in the approach I had to the H&R with my last team and equally as important I am hoping this next team will shed some light to statistical numbers regarding the usage of the strategy by HAL.

1st. The Park.

As it was discussed above I am going with SBC. Though it does limit the BP OBP to RHB it should provide the pitching with a slight chance of being better able to hold on to small leads, at least by supressing HRs and if the staff is predominantly RH it should limit Opp. OBP slighlty.

2nd. The Adjustments to H&R.

In my last team I ran the H&R in two ways. The first, which proved to be fairly inneficient, was to have hitters with a B H&R rating and good running lined up one after the other, the idea was to maximize the chances of different hitters triggering the next batter to an A rating and executing the play in different spots in the lineup. The problem with this is that I had no control over where would the play be executed. Sometimes hitters that are are better of swinging away were "sacrificed" and sometimes in the context of the lineup an unnecessary attempt happened given who was on deck and in the hole.

The second way isolated the H&R much more effectively to certain points in the LU and it seemed to have had some fairly positive results: the hits off H&R went up, the SB Pct. stayed ok and the successful moves was ok too. The problem is that the play was ran in such low numbers that it made difficult to evaluate whether or not the usage of the H&R pays dividends vs. high K pitchers, and as an offensive strategy in general in SOM-OL.

This last way is what I mentioned before: a lineup that is divided in three segments. Triggers, Hitters and Cleaners. Where the "triggers" get on base with good speed, trigger the "hitters" B rating to an A and then the "Cleaners" can score the guy from 2nd with a SI or from 3rd with a FO.

Now in deciding which way to go for this next team and going through the player pool, a problem came up with this approach. The good OBP hitters (Trigger candidates) are also usually good hitters, either way they're better off swinging away. The Cleaners usually are poor runners so they have a hard time also serving as triggers and some of the B H&Rers are too good to have them in that position and not good enough to have either Triggering or Cleaning. So, I adjusted the approach where the trigger and the cleaner are the same and the H&Rer has to be A) Pretty good defensively B) Pretty cheap and C) Have a card that doesn't pose any loss when using the H&R.

Given SBC the trigger/cleaner has to have: Ideally a RHB, a 400+ OBP, a * running, and a 300+ AVG.

The H&Rer must excell defensively, be cheap, have a B H&R rating, sub .250 AVG and sub .300 OBP should lower their pricetag and offer the possibility of a gain in their production by using the H&R.

3. The Team

The team is to take part in a non-DH league, which I believe is a better enviroment for the strategy. It will feature 3 Hit and Runners, 4 Trigger/Cleaners and Andrew Jones.

Andrew Jones doesn't really fit the team except for the very important reasons that his defense is quite good and for the reasons Lucky already pointed out defense should be good in this experiment. He is a RHB who should benfit slightly from SBC, his SLG provides an extra chance to have production and most importantly, in a BP that encourages baserunners his -4 arm will hopefully limit wild base running. Having said this the CF pool though rich is not particularily condusive to this exercise, other than Endy Chavez as a H&Rer the options are fairly limited and they all have negative sides, either poor running, poor defense, too expensive...and as luck would have it Endy Chavez was taken (again).

So the H&Rers are Neifi Perez (2B), John Olerud (1B) and Cesar Izturis (SS). Izturis is way too expensive a H&Rer but his main contribution is defensively. I am hoping that in the '06 set Jack Wilson will provide a much better alternative. Originally I wanted Matheny as the H&R Catcher, but given the limited pool of triggers/cleaners I opted for:

Kendall (C), Mora (3B), Berkman (LF) and Ichiro (RF).

One of the advantages of the lower salaries to the H&Rers and the absence of a DH is that it left more cash to be put into pitching, though I lack the dominating arms in my starting rotation I have a fairly solid bullpen anchored by Gordon and Calero and 4 5ths of my rotation are non *SP over 3 mil with my 5th starter being Greinke who should fair OK in SBC. As an added bonus this being a non DH league, most of the parks supress HRs and only 5 supress OBP significantly (3 Petcos, 1 Safeco and 1 Wrigley). I still wish I coul've netted a dominating *SP but well...maybe next time.

4. The plan.

What I am intending to do is through the makeup of the lineups, manager settings and individual players' settings restrict the bunting and stolen bases game. Only pitchers should bunt and the only SB/CS should be a byproduct of failed H&R plays, while at the same time only in three spots in the lineup should the H&R be attempted. After each series I should have a much easier chance of figuring out who and when tried to H&R vs. what pitcher/catcher combo and the outcome.

I am hoping that at the end of the season I'll have a better idea of how HAL uses the H&R, pitcher's sacs and the overall effectiveness of the strategy.

Will post some stats as they come around.

PostPosted: Wed Mar 01, 2006 4:42 am
by cummings2
Some early stats are starting to be compiled. Due to the nature of the process I feel reluctant to impose on anyone here by writing up a whole bunch of stats and ramblings that seem to me more appropiate for a Blog...problem is that I am not much of a blogger. So, I have found a very good use for my blog, it has become my notepad provided by TSN.

If, for whatever reason you're interested in seeing the stats go right ahead and check the blog, please keep in mind that its being written mostly for me to keep track of things so it might not be the best read.

PostPosted: Thu Mar 02, 2006 8:52 pm
by cummings2
I'm having a bit of a problem with the team. Even though it is early in the season (9 games) I am not being able to get HAL to execute the H&R as often as I'd like to.

Through individual and team setttings I was hoping HAL would run a H&R play 30-50% of the times that there is a H&Rer at bat with a *runner on 1B and less than two outs.

The situtation is coming up as often as planned, but HAL is just not H&Ring, it's being done less than 10% of the times.

Does anyone know the general parameters of the manageral settings for H&R? (i.e. Conservative, Normal, Aggressive) and what does it imply to check "hit and run more" on the players' card?

I was running it under the impression that the H&R more tag means that when the paramenters for the manger tendency (Normal, Very Aggressive, etc) are met the hitter will do the task more often.

The team is not responding to any change in manager or player settings.

PostPosted: Fri Mar 03, 2006 2:05 am
can't tell...but don't expect 30%, that's for sure. you seem to have some players who are not good runners, and I am pretty sure that Hal won't use H-R with poor runners.

PostPosted: Fri Mar 03, 2006 9:20 pm
by cummings2
Lucky! Thanks a million, your very accurate observation lead me to following a train of though that makes the numbers fit quite well. I'll observe these figures I'm coming up with but as it stands I think I've cracked the solution to this team's problem. 8)

PostPosted: Thu Mar 09, 2006 6:08 pm
by cummings2
Lucky, I have gathered enough info that leads me to believe that you are absolutely right. HAL does factor in the running ability of the man on base before deciding whether to execute the H&R.

At this point it seems like the most important factor is the runner's ability to get a safe lead, not the overall chances of stealing. This would help explain why Pierre is such a H&R inducing well as why Ichiro, though a good runner is not bound to induce the next batter into H&R.

[i:f6bfcde7be]IF[/i:f6bfcde7be] this is to have some degree of accuracy, then a good runner to induce H&R should be Eric Young, though not a great runner, he has surprisingly high chances of getting a safe lead.

The next step is in determining the thresholds or benchmarks that HAL uses to determine whether the H&R setting of Conservative, Normal, etc.. are met or not.

If we take the fact that the best Runners are in the mid 40 % to 50% range of getting a good lead and the average good ones that aren't likely to induce the play are slightly below 30% then I am going with the assumption that those settings range the between 30% and 50%. Either way, With Pierre's 2-12 he is always bound to be over whatever setting.

Thanks again Lucky...and congrats on the TSN Board award :wink:

PostPosted: Thu Mar 09, 2006 8:10 pm
Hey thanks LOL

PostPosted: Sat Mar 11, 2006 10:47 pm
by PossibleJohn
I took two shots at winning a Championship with stolen bases and the hit and run. [b:073045d1ba]Not successful.[/b:073045d1ba] Here were my two teams:

[b:073045d1ba]White Elephants, 73-89[/b:073045d1ba] (PETCO). Core: Pierre, Suzuki, Helton, CGuillen, Helton, Rodriguez, Figgins, Crawford, DRoberts. Produced 678 runs (10th in league).

TEAM G SB CS SB% SAC--ATT SQZ-ATT [b:073045d1ba]H&R--ATT-ADV[/b:073045d1ba]
WHI 162 437 198 .688 53 75 13 21 [b:073045d1ba]19 60 37[/b:073045d1ba]

[b:073045d1ba]Maxwell Demons, 75-87[/b:073045d1ba] (PETCO). Core: Pierre, Castillo, Casey, VMartinez, Figgins, Crawford, Izturis, Damon, DRoberts. Produced 677 runs (last in league).
TEAM G SB CS SB% SAC--ATT SQZ-ATT [b:073045d1ba]H&R--ATT-ADV[/b:073045d1ba]
DEM 162 414 162 .719 75 103 10 22 [b:073045d1ba]12 56 35[/b:073045d1ba]

PostPosted: Sat Mar 11, 2006 11:50 pm
by cummings2
hey there PJ, thanks for sharing this info here. It is quite intriguing, for several reasons including the fact that your teams played in a park that supresses both OBP and SLG and I have been in parks that encourage OBP.

If you don't mind sharing stuff here would you mind telling me how you formed your lineups and your teams' AVG, SLG, OBP, ABs and Runs against.

Very intriguing teams, you had the makings of really solid H&R teams. Right off the bat I see you had the same problem as I've been having with the bunting game.

Have to run now but I'll go through this later on. Thanks PJ!

PostPosted: Fri Mar 17, 2006 5:24 am
by cummings2
Oooopsy, PJ. Sorry. Hadn't realized both teams are in the record books (though I knew the names sounded familiar)

Actually I had already gone through both your teams in some detail in the process of gathering info for this H&R lab I'm into at the mo.

One of the things I find very interesting about both those teams is the pitching stats. Both Petco teams with an ERA over 4 and actually with good defense and good arms. Both offenses we're not particularly clunky in that their projected runs was about what they ended up scoring. Quite interesting.

I think you ran those teams quite well, though excellent runners, the Elephants would struggle a bit with the H&R since only Figgins is a good candidate to execute with consistent results.

With the Daemons however, Figgins, Castillo and Izturis would be solid candidates, unfortunately though I can't get to see your lineups with either team since the game stats have been erased.

If what I am concluding has some degree of accuracy, HAL's way of choosing which offensive strategy to use is sequential (as opposed to comparative) meaning that when a runner is on 1B HAL seems to go through the following:

1. Do I Steal (Y/N)
2 If N= Do I bunt (Y/N)
3 If N= Do I H&R (Y/N)
4 If N= Hit away.

I think that if you had ran those teams with conservative bunting you would've likely hit the 100 mark for H&R attempts, if you arrange the lineups in a H&R beneficial way and tag the right hitters with H&R more, maybe you could've gotten to the 150 mark. Your teams posted very impressive stealing numbers given Petco nonetheless.

As far as my current experiment, I am using two teams to cross reference information. Both teams have similar designs but one has much better runners. The trick is that with the better runners Stealing is actually a better options than the H&R -at least for me- so that works a bit against the H&R strategy experiment. However, I am finding ways of setting up lineups and settings so that even poor runners, in the right order will trigger the H&R when desired, actually the faster team (Rain Dogs in the blog links) has helped a lot in finding thresholds or benchmarks that HAL uses to determine whether to Steal, Bunt, H&R or hit away, when the info is cross referenced with the slower team.

In the main experiment (Carbone Quatorze -in the Blog links) I am averaging almost 1 H&R per game (42.85% of optimal H&R opportunities are executed) since finding the better combinations, as you can see in the links the team doesn't have particularily good runners, but in the right spot in the lineup and with the right managerial settings even they can trigger the execution of the play.

Well, the team is finally back at .500 since the correction in settings it has been running on a .619 winning pct. we'll have to wait and see.