College playbook project

[Deltadog]

NOTES RELATIVE TO SELECTED PLAYS

PROIRQBBOOT – Marked PASS - logic forces QB to run

PROIRCOUNTERPA – Marked RUN logic forces QB to pass - Observed a good many sacks this play by outside rush. Perhaps blocking backs should not move up so close to the line. Rush up the middle never is a threat on this play, but that outside rush killed it.

PROIRFBDIV – Marked pass – Logic shows it is a run play. R1 has no logic? Should he not have block nearest at end of route?

PROIRFBDIVW – Marked run – Logic shows run –

PROIRHBCOUNTER- Marked run – Logic shows run –

PROIRHBLEAD – Marked run - Logic shows run

PROIRHBSWING– Marked pass – Logic shows pass

PROIR-HBWEAK- Marked pass - Logic shows run

PROIRHBWEAKPA- Marked pass – Logic shows pass –

PROIR-OPTL – Marked pass – Logic shows run – TE has no block logic at end of route. Should he not have block nearest at end of route?

PROIRQUICKPA – Marked pass - Logic shows pass

PROIRSWINGPASS – Marked pass – Logic shows pass

PROIRTOSSSWEEP –Marked run – Logic shows run

PROIRVEEROPTW-Marked run – Logic shows run

The above plays have been examined. I have made notes relative to some of them. Those that show play name and marking and logic that agree are here for my reference only unless and additional note is included.

Several are Marked pass with run logic or vice versa. Did you do this on purpose and don’t you thing that it will confuse the computer when it is calling plays. For example you have a situation that demands a pass. (EXAMPLE 3RD AND 20) The computer calls PROIRQBBOOT because it is marked pass and instead gets what amounts to a QB sweep?

[nmleague]

Deltadog, thanks for the feed back. You bring up something I hadnt thought about. When I first started playing around with the PDS, creating a new play, I didnt notice the boxes to check for run or pass on screen where you pick the formaton to use for the play. And still many times I forget to check run or pass for the play type Im creating. I wasnt sure it had an impact, thought it just gave a quick start by assigning all OL pass blocking. But you bring up a good point, does the computer look at that when calling plays? It probably does since there is no profile set up for the play book. So when I have time today I will check all the offensive plays make sure that runs are indentifield as runs and passes as passes.

I will also check the plays mentioned were players have a logic missing, in come cases if the player wasnt involed in the play, or if I just wanted a R to clear an area I might not have given a final logic.

Will also see what I can do about the LB not reacting to RB quickly enough. As well might add 1-2 more defensive sets (run stoppers), and test changing the DL from a run rush to pass rush logic.

[Deft]

nmleague, do you find the computer loves to call the same plays over and over? I know with my playbook the computer over uses the Flood1.

[nmleague]

Deft Im not sure, it doesnt seem to. I just checked the game logs from two CPU vs CPU games using the college playbook. There seems to be a wide selection of plays used but a few favorites. But then again this playbook has about 26 plays, 52 possible if you count flipping, and the computer does flip plays. It would be great if the AI found plays that worked during a game and used them more often.

Im not looking forward to doing it, but I think that profiles really need to be set up for each playbook [:(]

[nmleague]

Update: I hope to have the next version done sometime tonight. I have gone through all the Twin formation plays and made sure that pass plays are classified as pass and run as run. Will next do the same for the ProI plays. Also will be doing some work on the defense, adding 3 or 4 special defenses and checking the logic on the current.

And Im going to create a second college playbook, with basically the same offense from the first one, but with a new defense replacing the 5-2. After that Im going to do two more playbooks, they will have a different offense replacing the current I, and one will use the 5-2 defense and the other the new defense.

[nmleague]

I just uploaded version 3 to FBmax, this will be the last version for awhile, want to have this tested well before making it final.

Changes from last version.

1.All offensive plays classified correctly, passes as passes, runs as runs.
2.Also tweeked a few of the plays and now all players should have logic at all points.
3. Made changes in all zone defenses, move the CB farther out and gave them a read command before the zone command. When tested the CB stayed outside better and reacted better to running plays. Also gave the LB in almost all plays a key on FB or HB command, and their movement to the ball seems better.

The work around for DB with m2m coverage still leaves alot to be desired, if they have m2m coverage on an outside REC and that REC has an out route, they get outside to late to cover it, hopefully David will fix this.

I think this version has addressed the known problems, Im going to start running CPU vs CPU games tonight. I have been pretty happy with the stats that result from this playbook, and it really is pretty cool to watch. The problem with scrambling QBs will proably remain, but when David changes that or if we have to manually change it, I think this will work pretty good.

Next Im going to create a playbook with the same offense as this first one, but with either a college 4-3 or a 4-4 defense. Any suggestions?

[garysorrell]

At least KY can win a game in MaxFB.

GO CATS!!!

Cant wait to try this stuff out.

[nmleague]

Below are the stats from my first CPU vs CPU game using version 3 of the college playbook. Again almost all the pass attempts were in the first half, QBs dont have all that many runs, but still to many. Does fatigue make them scramble more? I watched some of the game are things worked pretty good, maybe to many fumbles, Indiania completed a deep pass down the middle which the REC promptly fumble when hit on about the 2 yard line [:(]. This is from a league using college level players, next game will be teams with pro level players.

[nmleague]

Here are the stats for the second CPU vs CPU game using version 3 of the college playbook. Wow only one run by a QB, and look at the number of passes and yardage. This was using a league with pro level players.

[Deltadog]

It is 10:27 pm CDT and version 3 has not yet posted to FBMAX.

Not complaining, just letting you know why no one else has yet run a game using these plays.

[nmleague]

I noticed, Shaggy must be busy, I noticed that he was on earlier tonight, and I did send him a PM right after I uploaded the file.

I think you will like the latest version, your imput was very helpful, thanks.

[jcd38]

i'm new on this board, so greetings.

using the latest patch, i changed the QB Read Threshold in the Data Constants in the MainData file from 2.5 to 3.0 (This was 3.5 in a previous patch.) I played two games, sorry i didn't do any screen captures, but i noticed the QB scrambling decreased.
You may want to look into this yourself to confirm.

[nmleague]

jcd38 Ive seen your post on a different thread about the change you made, sounds like a good move and will probably be a good way to cut down on scrambling. I think David will probably make a similar change in the next patch, so I have held off from doing what you did, if he doesnt make the change than I will have some questions on how you did it.

[firebirds]

nmleague, I think you just confirmed something I've been thinking ever since the first patches started coming out. Correct me where my thinking may be wrong if you would please.

You can go and adjust the attributes/abilities of each individual player. But in all reality this is fine tuning because the attributes/abilities are derived from a master profile based upon position. I'm sure this is overly simplistic and my terminology will be wrong just as the numbers I'm tossing out, but for the sake of example/clarification please bare with me. What I'm thinking, let's say the very best wide open receiver, with a good throw, will catch the ball 95% of the time. So the master profile for that position sets "hands" at a 100 setting will catch the ball 95% of the time. Now you have a receiver on your team with "hands" set at 85. I'm guessing that he will catch the ball 85% of the ultimate 95% possible. Another example, the master profile for an OL would have "hands" set at something far lower, oh maybe 35. So an OL on your team with 40 "Hands" will only catch the ball 40% of the ultimate 35% possible.

So to follow thru, a coach is unhappy with the number of sacks the QB is taking because he won't scramble. Using a trial and error process the coach finally adjusts the numbers to get the results he wants (maybe just the QB, maybe OL and DL also). A patch comes out where one of the changes was altering the master profile for the QB position that enables the QB to scramble more. So what happens now is that the QB player scrambles so much now that he is rarely passing. The reason being the coaches player adjustments magnify the master profile. Don't know if any of that makes sense. I'm just trying to better understand how the game works. It appears to me that with MaxFB being so adjustable some of the things happening and/or not happening are a result of gamers adjustments which compound/magnify the requested fix for a particular problem.

[nmleague]

Firebirds I think you are correct that some of the changes David is making are to a master file. And if he is making changes to a master and the owners of the game have been making changes to the player skill ratings, that yes they could work at cross purposes or exagerate some changes. Once David releases a final patch, with no more changes to the master, then owners of the game will be able to fine tune the game by making changes. Thats why Im not making any changes manually to the QBs scrambling threshold on my own at this time. If when David has released a final patch, the QBs are still scrambling to much I might either make the changes mentioned by JCD above, or I would edit the skill range for QBs, raising the lower level of the Pocket passer rating.

So far in my game the only things that I have edited have been the speed skill range for players, as mentioned in several threads about a month ago. I was happy with the results of that, but to tell the truth I havent checked to see if the patches that have come out since I did that, have set the players skill ranges back to the original.

[nmleague]

The computer just finnished another CPU vs CPU test game, stats below. This was a first for me, no passes thrown by either team. This game was by two teams with college level players. I wondered why no passes and a total of 28 rushes by both QBs. So I checked their pocket passer ratings, and then check those compared to the QBs in the previous two games played. Here is the comparison

Michigan State QB Corella 13 rushes 0 passes PPR 62
Penn State QB Stollar 15 rushes 0 passes PPR 59

Northerwestern QB Batler 4 rushes 12 passes PPR 79
Indiania QB Dehm 9 rushes 6 passes PPR 74

USC QB Satchwell rushes 1 passes 17 PPR 48
UCLA QB Garriss rushes 0 passes 37 PPR 74

The first 4 are college level players, the last two are pro level players. Easy to see why the MICH St and PENN ST. qbs ran so much and passed so little. The surprise is the USC QB, such a low PPR but he only rushed 1 time, but he did throw a good number of passes and his team was having so much success running that he didnt have to.

[Deltadog]

This message is relative to the discussion above between Firebird and NMleague in message 54 and 55.

In your discussion you seem to be discussing the wisdom of adjusting the pocket passing attribute of specific quarterbacks and the effect that this might have if and when David adjust the masterfile. In the prior thread, the writer said, "i changed the QB Threshold from 2.5 to 3 in MainData.mdb
this seemed to stop the QBs from scrambling excessively and resulted in them passing more. i haven't tested this to great length but it seems to have a significant effect.
Question for Dave,

What does the QB Threshold do? I noticed in this latest patch the QB Threshold was lowered from 3.5 to 2.5
Also, when you get some time, could you provide some more explanation on the Data Constants? "

Note that he made a global change to the Maindata.mdb file, which if I understand it correctly is the masterfile. If I am correct then he has in fact made a change that will affect all QB in the game and which may well be the change David might make in a future patch. In any event I suspect that each time a patch is installed the MainData.mdb file will be overwritten and the changes David makes will superceed any changes the user makes. If the user does not like David's change he can adjust this setting again.

I might be wrong, but this is my understanding of the change that jcd38 has made.

[jcd38]

deltadog,

your thinking is the same as mine on this. the MainData file, and i think i've seen Dave referencing this in another thread, is similiar in nature to what you find in other games as tune or ini files. The DataConstants in the MainData file from what i understand are global. this QB Read Threshold attribute looks to me like a global attribute that affects all QB's. only Dave can answer that for sure. And if it's global, that means to me, that this should probably be fined tumed (or any other coding in conjunction )before taking into account other things such as player ratings, effects of particluar plays, or professional vs college levels. The logic of the QB Read Threshold might be a determing factor in when the QB "decides" to scramble or pass or check out the cheerleaders [8D]I'd like Dave to clarify this for us. Because we could be spending a lot of time trying to fix the scrambling thing when Dave already is,etc.

[Shaggyra]

Just to let you know. Version 3 is now available on FBMAX.COM

Life got in the way of geting up faster.[:(]

[nmleague]

Just a quick note, I changed my mind about the second college playbook. Instead of taking the offense from the first one and adding a new defense, I have done the opposite. The new playbook will have the same 52 defense, but now with a wishbone offense. The offense will be somewhat limited, about 10 running plays (20 flipped), and limited pass plays. Might have to add a different formation, broken bone, for passing situations. The first version will be available either tonight or tomorrow.

Im allready finding that Im going to have to add a few more run stopper, run blitz defenses, to the 52.