Starting to feel like a real Football game

[Deft]

Though the first month of owning FBMax felt like I was still beta testing. The latest version feels much more like a finished product.

I have a playbook and series of plays that work and can move the ball without excess trickery. There are still some very important issues to be addressed but I can make it through an entire game and it finally feels like football. What are your impressions at this point?

[Lucas718]

I agree. It's starting to come together nicely. I'm hoping by the time football season really starts we'll have a bunch more playbooks, real rosters, and hopefully some stadiums.

[kshinner]

I'm not yet convinced. I've still seen linebackers run right into the blocking instead of sidestepping to fill the hole. I see ball carriers streaking straight downfield, oblivious to the saftey coming over when there are acres of room on the outside to gain more yardage. I've seen the RB go off-tackle with the endzone in sight just to take a left turn into the middle of the field and get tackled.

I'm not asking for perfect but I am asking for logical. Right now, it hasn't met expectations.

[jcd38]

i'm starting to see a more "realistic" feel to the play. i've been tweaking the maindata file and the default playbooks; in an attempt to get it as realistic as possible. it's not where i'd like it to be yet; but it's getting there. it will take some more work. my lastest version of the MainData file (V5) with my default playbooks seem to be producing more "realistic" play; things like players acting more life-like.
some things i would like to see:
1) the ability to change the draft order when creating a league.
2) some of the bugs already mentioned fixed with the play editor.
3) stat keeping bugs fixed.

[JonnieR]

Using actual ratings for real NFL players (circa 1994-95) I have had much, much better results with the latest patch. The scores are realistic and the individual players perform pretty consistently with their abilities.

Having said that there are some things that the game needs:

1. Better selection of plays and more playbooks for the US professional level game!!!!!!
2. Improved Stats recording and ease of review options.
3. Some benchmarks for the attribute ratings of NFL level players. I have been using the ratings relative to my old FBPRO files, however not everything seems quite right. For instance, if the "avoid block" attribute is to high the defense runs wild leading to an inordinate number of sacks and stuffs. I have lowered this attribute by 5% across the board and things improved a tiny bit, I then lowered it another 5% and now it seems better, but the QBs still seem really hassled when going back to pass. Someone else mentioned linebackers always avoiding their blocks, he should check the block avoid attribute ratings.

[jcd38]

for my own benchmarks, i based (or at least started with) my player ratings off a math formula that states when creating/using any type of rating scale that the low end of the scale or the high end of the scale would usually not deviate more than 30% from the mean (i.e. average.) say the average rating is 75; 45 would be the low and the high would be 105..so using 25% deviation you wind up with a scale of 50 low 75 average and 99 high). most games ratings follow this basic concept. so the old football pro ratings, madden, all should pretty easy to translate to maxfb. i noticed that when you go below 40 or 45 with ratings things start to or may become "un-realistic".

[Deft]

I am not looking to deeply into the statistics and career features until the in-game piece is solid.

Here are some of my gotta fixes-

break the huddle defense lineup issue
defenders with run rush or pass rush that stand there when its the opposite call
defenders still overplay the primary WRs and underplay secondary receivers, TEs, and RBs
snapping the ball before the MIM reaches his destination
the discrepency between the PDS location and the in-game location
short passes and timing between QBs and receivers

Here are some of my like to fix-

Make lob passes have higher trajectory
Give the option for different special teams calls in the offense and defensive profile
Add onside kick logic

[Magnum357]

ORIGINAL: jcd38

for my own benchmarks, i based (or at least started with) my player ratings off a math formula that states when creating/using any type of rating scale that the low end of the scale or the high end of the scale would usually not deviate more than 30% from the mean (i.e. average.) say the average rating is 75; 45 would be the low and the high would be 105..so using 25% deviation you wind up with a scale of 50 low 75 average and 99 high). most games ratings follow this basic concept. so the old football pro ratings, madden, all should pretty easy to translate to maxfb. i noticed that when you go below 40 or 45 with ratings things start to or may become "un-realistic".

Yes, I totally agree with this. I haven't done it yet (got to find the Ranges in the MainData.mdb file to do this) but I'm planning to make the Rating Ranges at a similar Range (40 Low, 70 Med, 99 Max). I have done this with other Football games and that is about right for a General Attribute Range. This is also good if say you are running a College league or something where not all teams are going to be power houses in the league.

[Deltadog]

My pet peeve with the game so far has to be the punts and KO that slice down the field toward the right coffin corner 85 to 90% of the time.

[JonnieR]

With regard to generating realistic attribute ratings:

1. I develop "scores" for each attribute that reflect the actual performance of the player at any particular position for a particular season. These scores are based upon the original models used in FBPRO, but also reflect some Madden influence.

2. I then calculate a Std dev and mean for each score based on the stats for the entire league at each position (i.e., Strength rating for LBs, speed for WRs, etc.)

3. I have found that a Normal Distribution normally fits the data the best and calculate the players percentage rating based on the density function of the distribution.

4. Depending upon the ranges for each of the various attributes (i.e., WR Speed, 68 -99, etc) I then calculate a rating for the player. For instance, a WR at the 90th percentile for Speed would have a SPEED rating of 95.9 (96).

This modeling is sensitive to the ranges used for the attributes by position and the models derived from the player's stats and I am always tinkering with these models.

I Have full stats for 1995, 1998, 1999, 2000, and 2002 so I could theoretically develop all the rosters for each team in the NFL for these years for MAXFB. Theoretically is the key word here.

[Deft]

Somehow I don't think CFL_Fan_7 is going to be responding to this thread. [;)]

[jcd38]

developing some kind of standard like the FBPro model (VPNFL, etc.) for ratings probably would be good for the community.

[Magnum357]

Ya, I agree. VPNFL was an excellent system for Football Pro. Unfortunetly, I've heard many H2H Players say they don't use the VPNFL system. They claim that the Players are too slow and don't like the limitations of certain Ratings. I never could understand why they don't like using the VPNFL system, I've played many Coach mode games with VPNFL and I think its would work with H2H play too, but not H2H league on the Net uses it. Now a lot of SIM type Football Pro leagues use it and its definitly the system of choice for those kind of leagues.

I'm not sure the VPNFL Rating system would work well with MaxFB. There were a few Ratings in the system designed purposely a certain way too work around certain limitations and bugs Football Pro has. For example, almost all the Defenders in the system have a slight Speed reduction compared to the Offensive players. This was done because Pass Coverages where easy to defend due to limitations of Play Editor. Also, Defensive Linemen were limited in speed because a couple bugs exist in Fbpro98 that limit the capabilites of Offensive Linement to block effectively. Limiting the speed of Defenders was about the only real solution to work around these problems. Also, almost all Kickers in the system have Kick Ratings shot up to the moon because it was tough to get Kicking Stats like the NFL. Also, since QB Strength Rating is tied in with how well he Throws passes, this has a nasty side effect of a Defensive End coming in at full speed but easily bouncing off the QB like he was nothing because he has high Strength. Why Football Pro never had a Seperate Pass Strength and Pass Accuracy Ratings is beyound me.

But, the VPNFL system does try to use Real World Physics to get the system to "Feel" a lot like the real thing. Like analyizing 40 yard speed times in the Real NFL and testing how equivelent the Ratings would work in the game has really resulted in very Accrate game play. I agree, a lot of the theories of the VPNFL system could be used to make something similar in MaxFB, but I wouldn't just completely copy the same Ratings System.

[frunky]

What if you made the low go to 50 to avoid the below 45 problem
will that mess things up?

[JonnieR]

Magnum is correct about the odd ratings in FBPRO, the upshot of the FBPRO system was that difference attributes meant different things for players depending upon their positon. For instance, Offensive lineman, STR indicated run blocking ability and AGL indicated Pass Blocking. STR for QBs only referred to arm strength, etc.

When I say I use FBPRO models, I am talking about developing "scores" for each player for each relevant attribute. For instance WR SPD "score" would be a function of his avg. yds per catch, YAC (yards after catch) and longest pass play. These factors would be compared to an average and a "best score".