Applying SIM data

[StalkingButler]

Yep, can't figure this thing out.. here's the rom if anyone wants to give it a glance and try for themselves. I'm getting frustrated here, wanted this out for the Super Bowl and the sim data is really tripping me up. I'm claiming a formation bug. I went into the original and made the Washington/San Diego/Phoenix TEs much better sim stat wise, and it didn't help them out very much stat wise, maybe 5 catches.

As far as the formations I used, I've also tried Heath with 1-4 routes of all different kinds in the playbook to little avail.

TSB 2009 StalkingButler's Version (TE stat problem).zip

[Lhopital]

I looked at your rom, and here's what I found:

RB1 Parker [7, 1, 4, 6 ]

RB2 Washington [1, 5, 4, 9 ]

WR1 Ward [1, 11, 7, 7 ]

WR2 Holmes [1, 8, 11, 9 ]

TE1 Miller [1, 7, 6, 6 ]

With your SIM set the way it is, the 4th numbers add up to (6+9+7+9+6) 37 or (48+72+56+72+48) 296. You only have 255 to work with; or 32 SIM points with the adjustments at x18117. If you do a simple stat conversion for catches, you get:

3+40+81+55+48 = 227 catches for starters, and ...

Parker 3/227 x 32 = 0.42

Washington 40/227 x 32 = 5.64

Ward 81/227 x 32 = 11.42

Holmes 55/227 x 32 = 7.75

Miller 48/227 x 32 = 6.77

As another thought - I've seen a couple ideas on how to adjust the third number (controls receiving yardage). For my SIM values, I take the yds/catch and subtract 4. It's not perfect, but it helps me scale yardage output somewhat. For the steelers:

Parker 4.33 - 4 = 0.33

Washington 15.78 - 4 = 11.78

Ward 12.88 - 4 = 8.88

Holmes 14.93 - 4 = 10.93

Miller 10.71 - 4 = 6.71

If you round down (except for 0) for target values and round for rec yardage I get this:

RB1 Parker [7, 1, 1, 1 ]

RB2 Washington [1, 5, 12, 5 ]

WR1 Ward [1, 11, 9, 11 ]

WR2 Holmes [1, 8, 11, 7 ]

TE1 Miller [1, 7, 7, 6 ]

This gives you a target sum of 1+5+11+7+6 = 30 and the remaining 2/32 is randomized. Try it now, and see if it simulates a little more realistically.

[Lhopital]

I think I might have tried to correct the wrong problem.

RB1 Parker [7, 1, 4, 6 ]

RB2 Washington [1, 5, 4, 9 ]

WR1 Ward [1, 11, 7, 7 ]

WR2 Holmes [1, 8, 11, 9 ]

TE1 Miller [1, 7, 6, 6 ]

It looks like you used the 3rd numbers to get a sum of 32. It's the 4th number that gets used as the target value. Sorry for the lengthy response. Good luck!

[StalkingButler]

FANTASTIC! It works and I understand now. Thanks for your time.. now I gotta crack down and finish these sim numbers by Sunday.

[BO FB Offtackle Left]

just found this thread and I'm still a little confused because this guy kept talking about adjustments he made to his rom and errors he made so I have no idea what the right conclusion is.

Working on an unedited rom, I know the defense total is 255. What is the total for the 4th sim number on offense? Is it actually 32, or is it something else?

[buck]

see:

viewtopic.php?f=5&t=11924&start=15#p94632

Basically it's 32, but if you read what Lhopital wrote above, he found that 0-15 actually pull up values from the table he lists. So you can modify the table like he says to make the numbers even.

There is another mod you might make - he mentions it it part 2. It has to do with the TE getting part of the random receiving targets if you don't fill up the 32 SIM point total.

It's cool that they made the game where some teams are pretty much set in who will get the receptions, but then there are some teams that have a lot of randomization going on with SIM receiving targets over the course of a season.

Working on an unedited rom, I know the defense total is 255. What is the total for the 4th sim number on offense? Is it actually 32, or is it something else?

[BO FB Offtackle Left]

see:

http://tecmobowl.org/forum//viewtopic.php?f ... =15#p94632

Basically it's 32, but if you read what Lhopital wrote above, he found that 0-15 actually pull up values from the table he lists. So you can modify the table like he says to make the numbers even.

There is another mod you might make - he mentions it it part 2. It has to do with the TE getting part of the random receiving targets if you don't fill up the 32 SIM point total.

It's cool that they made the game where some teams are pretty much set in who will get the receptions, but then there are some teams that have a lot of randomization going on with SIM receiving targets over the course of a season.

Working on an unedited rom, I know the defense total is 255. What is the total for the 4th sim number on offense? Is it actually 32, or is it something else?

Well, the TE not getting passes drove me crazy in my other roms, but I don't want to do any hex editing for this rom. So it's good to know that as long as I total 32 that everyone should get what they are supposed to. I have no interest in randomizing catches. There's enough random shit going on in Tecmo as it is.

I still don't understand why you would bother modifying the table when you can just as easily assign values from 0-15 in Tool. From what I understand, the 4th sim value is the number of times the player is targeted, the second is actual receptions, and the third is yards. Or something close to that. PFF actually has the number of times each player was targeted, so wouldn't it make sense to base the 4th sim on that? So a player that was targeted a lot but didn't make a lot of actual receptions would have a high 4th value but low second and third values.

So for example, the Saints might look something like this? Sim 4 = TA Player * 32/Total Targets

Pos	Name	        Rec.	Yds	TA	Sim 1	Sim 2	Sim 3	Sim 4

WR  M. Colston 70 1074 102 1 9 10 7

WR  D. Henderson 51 804 82 1 6 8 5

WR  Robert Meachem 45 725 61 1 5 7 4

WR  Lance Moore 13 146 16 1 2 1 1

TE  Jeremy Shockey 48 569 65 1 6 6 4

TE  Dave Thomas 35 356 46 1 4 4 3

RB  Reggie Bush 47 335 63 5 4 3 4

RB  Pierre Thomas 39 302 42 6 3 3 2

RB  Mike Bell 4 12 4 3 1 1 1

FB  Heath Evans 10 70 12 2 1 1 1

493 32

[BO FB Offtackle Left]

I just realized I counted the backups in my last post, so nevermind that. But I'm still not sure what the difference is between #2 and #4 below:

RB/WR/TE SIM:

Rushing Ability, Receiving Ability, Receiving Yardage, Receiving Target

I was thinking actual targets, but it doesn't seem like targets without considering actual receptions is something that would be reflected in stats. Unless it also affects QB comp %? That is, if you make a player have a 15 in #4 but a 1 in number 2 and 3, would that make the QB have a very low comp % because he keeps throwing to a receiver who keeps dropping the ball?

If it's not that complicated and "Receiving Target" really just translates into the number of receptions a player gets when simming, then how is that different from "Receiving Ability"?

[buck]

jstout says:

"Passing and Receiving Ability work together in deciding a reception, incompletion, or interception on the play."

end quote

The difference is that "Receiving Target" determines if the WR is even targeted.

IF he is 'targeted', then "Receiving Ability" determines the outcome of said 'target' (reception, incompletion, interception).

[BO FB Offtackle Left]

jstout says:

"Passing and Receiving Ability work together in deciding a reception, incompletion, or interception on the play."

end quote

The difference is that "Receiving Target" determines if the WR is even targeted.

IF he is 'targeted', then "Receiving Ability" determines the outcome of said 'target' (reception, incompletion, interception).

Okay, so what I initially said is correct, then? To give an extreme example: if you give a player on one team a 15 in the 4th sim value and a 1 in the 2nd, and a player on another team the reverse, then the player with the 15 in the 2nd will get more receptions. If the QBs and team sim stats of these two teams are rated the same, then the QB on the first team should also have a much lower comp % because of all the targets going to that WR. Is this right?

[buck]

I've never experimented with an extreme case like that (or paid so much detailed attention to my simming results - as long as stuff is in the ballpark), but what you said makes sense.

One thing I'm curious about - how much the 'team defense rating' (0-F) of opponents affects offensive sim stat results.

I think it's time you go get an emulator with a "speed up" option and start simming seasons.

To give an extreme example: if you give a player on one team a 15 in the 4th sim value and a 1 in the 2nd, and a player on another team the reverse, then the player with the 15 in the 2nd will get more receptions. If the QBs and team sim stats of these two teams are rated the same, then the QB on the first team should also have a much lower comp % because of all the targets going to that WR. Is this right?

[BO FB Offtackle Left]

I've never experimented with an extreme case like that (or paid so much detailed attention to my simming results - as long as stuff is in the ballpark), but what you said makes sense.

One thing I'm curious about - how much the 'team defense rating' (0-F) of opponents affects offensive sim stat results.

I think it's time you go get an emulator with a "speed up" option and start simming seasons.

To give an extreme example: if you give a player on one team a 15 in the 4th sim value and a 1 in the 2nd, and a player on another team the reverse, then the player with the 15 in the 2nd will get more receptions. If the QBs and team sim stats of these two teams are rated the same, then the QB on the first team should also have a much lower comp % because of all the targets going to that WR. Is this right?

yeah i did that once years ago but i think i lost that notebook. Plus I had thought I figured Rec and KR but I came to different conclusions than are on here, and my results for PR value were inconclusive. So I'd rather follow the results of people that seem to know more about this than I do, hehe.

[El Rey]

Ok, now I feel silly. It DOES work like the defense (out of 255) except that you have to translate to a 15 point scale. If I take another look at the 91 Bills starters (8,2,6,10,6) that translates into (64,25,51,76,51) which is 267. IND (2,10,10,10,2) is (25,76,76,76,25) - a sum of 253 before TE Pat Beach is reached (that explains why he catches about 1 pass every 3 seasons). Does this mean that the original programmers still goofed?

If the sum was to be "x" (a # less than 255), then (255-x)/255 of the time there should be a remainder of passes to be randomly distributed. Look at KC starters (0,0,6,10,0) = (0,0,51,76,0) with a sum of 127. This should mean that 127/255 of the passes stop at the two WR and (255-127)/255 are randomly distributed, right? But the TE still gets nothing.

I suppose if I wanted to make mathematically even distribution, I would have to rate RB1 to stop the random # 1/5 of the time, RB2 1/4 of the remaining time, WR1 1/3 of what's left from that, and so on. Here's what I get: (51/255, 51/204, 51/153, 51/102, 51/51) which is my original test #s (6,6,6,6,6). It's that leftover randomness that still puzzles me.

How does (8,2,6,10,6) translate to (64,25,51,76,51)?

[TecmoBowlSwede]

There must be more to the 3rd sim value than just deciding yards/reception, right?

 

I have simmed a couple of seasons with every player set (using TSBTool Supreme) to 0, except the Oilers starting WR's; 15 sim value for them. The yards/reception ranking is still very even, season after season. And often one of them doesn't even get in the Top 30.

 

...or is it just that random?

[Yak]

If memory serves, the type of pass play utilized affects the SIM stats, like when giving RB1 all four run slot options--his end-of-season numbers are going to be substantially increased because of it. Each pass play has variables in place for what receiver position the QB looks for more than the others.

 

For more about QB / WR priority: http://tecmobowl.org/topic/11317-changing-qb-wr-priority/

[TheRaja]

Does this mean that TSBSupreme has been calculating the SIM data wrong???

[Tecmonster]

For my money, the "Auto Update Sim Values" button in TSBToolSupreme does a good enough job in delivering expected stats for com players.  I personally don't see what all the fuss is about over calculating sim values.  Unless I've completely misunderstood the concept, it has nothing to do with man-controlled teams or how the computer plays against man-controlled teams.

[buck]

For my money, the "Auto Update Sim Values" button in TSBToolSupreme does a good enough job in delivering expected stats for com players.  I personally don't see what all the fuss is about over calculating sim values.  Unless I've completely misunderstood the concept, it has nothing to do with man-controlled teams or how the computer plays against man-controlled teams.

 

right, it's for the SKiPped COM games, when COM plays COM and you just see the scoreboard as the week goes by.  

 

when playing MAN/COA/COM v COM (non-skipped), the only variable that does anything (besides player ratings and "juice") is offensive preference.

 

but the fuss is about making the season "end properly" (where the best teams have the best record) and where Jerry Rice gets all the receiving yards and not Harry Sydney and Reggie gets the sacks, not Andre Waters.  

without proper sim values, you cannot control how those SKP games actually go.  and of course there is always a lot of randomness going on, so it's impossible to "guarantee" results.

Edited July 11, 2013 by buck

[Tecmonster]

Well, I do get that -- the best teams should go to the playoffs and the stat leaderboards should make sense.  I guess I should have clarified my question though: is the TSBToolSupreme's "Auto Update Sim Values..." feature so wildly inept that it shouldn't be used?  I've used it for custom leagues and players I've created and it seems to do a good job simulating team and player performances.  I mean, you will see some strange things from time to time, but the for the time it takes to set sim values with this feature (the one-second click of a button), the return on investment is solid, in my opinion.

[buck]

Well, I do get that -- the best teams should go to the playoffs and the stat leaderboards should make sense.  I guess I should have clarified my question though: is the TSBToolSupreme's "Auto Update Sim Values..." feature so wildly inept that it shouldn't be used?  I've used it for custom leagues and players I've created and it seems to do a good job simulating team and player performances.  I mean, you will see some strange things from time to time, but the for the time it takes to set sim values with this feature (the one-second click of a button), the return on investment is solid, in my opinion.

 

I've never used that feature!  I remember reading somewhere where BAD_AL said it didn't work right - and I've always had fun doing it myself with spreadsheets and testing - but hey, if it works, it works.  I might fool around with it.

 

edit - actually, the source code is out there - it should be possible to see the exact algorithm that BAD_AL used.

Edited July 12, 2013 by buck

[Martin]

Anyone got the locations for the other corresponding values, like the ones at x18117? For yrds/rush, etc.

[bruddog]

Rushing ability and receiving yardage load from the same array so it might be best to think of it

 

as

 

rushing yardage, reception skill, receiving yardage, target likelihood

[buck]

Rushing ability and receiving yardage load from the same array so it might be best to think of it

 

as

 

rushing yardage, reception skill, receiving yardage, target likelihood

 

wtf does "skill" mean?

[Tecmonster]

what I don't understand is that reception skill and target likelihood lead to the bottom line known as receiving yardage, yet receiving yardage also has its own sim number.  I guess what I'm asking is that once you enter a # for receiving yardage, how would reception skill and target likelihood matter at all?

[bruddog]

Reception skill helps determine if a pass is caught, not caught, intercepted

Target liklihood helps determine if a WR is targeted on a play.

If a player is targeted and the ball is caught the receiving yardage rating helps set the random range of yards gained on the play.