human vs cpu same teams

[09flhrc]

My "idea" was to try and get a baseline set of numbers where the teams were playing evenly over multiple games. Idelaly I'd win 50% and lose 50%. I was then going to try and adjust proportionaletly to attain "realistic results".

I've still been messing around with this concept and the CPU numbers have gone down (about 15% or so) and I'm starting to use multiple teams against each other to see if the slider settings provide a decent chaleng across more than one team. So far...so good...

Just my $.02.

[09flhrc]

I should really check my spelling before submitting a reply.

[jack58]

Quote: Originally Posted by 09flhrc

My "idea" was to try and get a baseline set of numbers where the teams were playing evenly over multiple games. Idelaly I'd win 50% and lose 50%. I was then going to try and adjust proportionaletly to attain "realistic results". I've still been messing around with this concept and the CPU numbers have gone down (about 15% or so) and I'm starting to use multiple teams against each other to see if the slider settings provide a decent chaleng across more than one team. So far...so good... Just my $.02.

Once you get a base set post it on page 1 and I'll try and give it a test when I get a chance to play. That way you could end up with several people testing using different teams which should hasten the process and make it more accurate. Sounds like a good idea and I look forward to seeing your results!

FYI ... you can always update posts if you want to correct spelling, etc

[dfos81]

Now I'm really seeing what's happening while testing in Play now. I decided to test Skins vs. Skins and 1st gm was set to All-Madden then I dropped the CPU from 50 to 39 on all skills, and increased the player sliders to 27.

Now fact is sliders are just as effective in Play now games as they are in Franchise. Now franchise brings into play other factors, coaching boost, strategies, dpp, etc.

1st game test Play now All-Madden, they blew me away like 31-0 @ halftime! Lol
Which this is what the sliders showed on All-Madden

25/50 on all player/CPU skills

So just using a calculation of how much a boost the CPU was actually getting, and reducing that boost apx 12% over CPU & player sides. Here's what the sliders showed, and you'll be impressed. W/ what a 12% decrease in CPU boost from 25/50.

Here's the 2nd test 12% less boost from All-Madden.
Oh yeah, I also put Qba like so and brk tckl bc that was not the players rating causing Grossman to throw 15yds off target, w/ feet set, and its a 5 yd drag lol. Or my Rbs looking like a girl that got smacked by a 250-300lb UFO moving @ the speed of light! Lol
So once I got out of the Player ratings based lowend slider concepts for both pos. Then I finally felt like Grossman was producing his ratings on the field. And Torain & Hightower started looking like Rbs and not lil girls being smacked by a UFO, moving @ lightspeed.

2nd slider set

50/50
27/39
27/39

50/50
27/39
27/39

27/39
27/39
27/39

Half was much more competitve
7-6 I'm up 1 but this is what's sweet to me
@ halftime
Offense

Tot O: 89/94
Rush: 25/39
Pass: 64/55
1st dwn: 6/4
Pr yds: 23/42
Kr yds: 0/32
Tot yds: 112/168
Turnovr: 2/2
3rd dwn: 2-9/ 1-8
Rdzn: 1-1/1-2
Penalty: 4-27/1-10
Home team was only penalized 1 once, can you say homefield advantage?

Anyway what's so intriguing to me is by dropping the boost from All-Madden 12% I'm seeing very, very similar results.
Now both sides still need a bit better run game, and maybe the player skills need a bit more on ints, keep seeing ez drops and 0 ints for the user side. CPU is getting ints apx 1-2 per gm w/ latest test sliders above, 2nd game test.

Leading taklers Mlbs, rolbs, all lbs! Top 5 tak! Another intriguing stat that is producing more realism.

[09flhrc]

My very first test was to set all CPU values to 100 (except fumbles) and all Human values to 0. I expected to get blown out and I got exactly what I expected! It looked like a High School team playing an NFL team. This led me to believe that if the CPU can really kick my *** then the sliders should be adjustable enough to get me what I want.

From what I'm seeing I would think that your slider settings should provide for a pretty good competetive game when using the same teams competing against each other.

[dfos81]

Quote: Originally Posted by 09flhrc

My very first test was to set all CPU values to 100 (except fumbles) and all Human values to 0. I expected to get blown out and I got exactly what I expected! It looked like a High School team playing an NFL team. This led me to believe that if the CPU can really kick my *** then the sliders should be adjustable enough to get me what I want. From what I'm seeing I would think that your slider settings should provide for a pretty good competetive game when using the same teams competing against each other.

Well, it was equal as far as stats player vs. cpu but The Rbs are avg 2ypc or less and the Qbs are struggling to throw a decent % 55-65% is what Im shooting for.
So I will give both player & cpu a running game boost
and wont touch the qba yet b/c w/ the run game, it should open the pass a bit more and hopefully the defenses wont be as stout as what I posted above w/ those sliders.
I will have game stats posted in the morning and the sliders that I used to get those #s up.

This is much simpler/effective than the player ratings sliders that we were testing. Also going to increase int slider for player side 1 click and hope to see at least 1 int for my team.

[dfos81]

This was 3rd game test(1st half) w/ the minor tweaks that I was talking bout earlier. This was exactly what I was looking for improving the run games for both User/CPU, increase int, and I was again, using some small increase to see if I noticed a difference, and yes I did!

User/CPU
Rex Grossman: rating 77.3/50,
13-19/9-18, 183yds/127yds, 68%/50%, 0tds for both, 1int for both.
9.6avg/7.0avg
31Lng/41Lng
Both Qbs played very much alike, CPU needing a bit better %, besides that they were like mirror images.
I really am impressed w/ the CPU decision making in the pass game! Short, Med., Lng, throwaways. Wow! I was playing myself, almost, but not quite.

Lil better rush avg. 3.9/3.0- still need a lil CPU boost(apx. 4-6%)
Penalties: 1/2(1st half)

Both sides forced a fumble, and this was also in similar fashion. S. Moss running over the Middle on both teams and both Laron Landrys SMACK! FUMBLE. The CPU got me first and then, bam I was able to get a bead on Moss running over the middle. It was almost back to back.
2 forced fumbles 1 on each side and all within 3 plays lol.

Sacks user/cpu
1-2
Also each side w/ an int! Sweet!

This was the set used

50/50
27/39
27/39

50/50
32/44
27/39

27/39
28/39
27/39

27/39
27/39
27/39

I really think just a minor boost to Rblk 3-4% for CPU and 3-6% increase to CPU wr cth, and I bet you will have a really great base sliders(BTW these are considered Allpro bc I can tell by kickmeter speed.)

I'm going to test abt 3 more and see the results

Oh yeah, score @ halftime in this gm. User/CPU-9-7, I got a last sec. Fg to get the lead bck. 3fgs. I'm sure it would open up a bit more in 2nd half but I don't want fatigue to skew the slider testing.
Good stuff!

[jack58]

Quote: Originally Posted by dfos81

This was 3rd game test(1st half) w/ the minor tweaks that I was talking bout earlier. This was exactly what I was looking for improving the run games for both User/CPU, increase int, and I was again, using some small increase to see if I noticed a difference, and yes I did! User/CPU Rex Grossman: rating 77.3/50, 13-19/9-18, 183yds/127yds, 68%/50%, 0tds for both, 1int for both. 9.6avg/7.0avg 31Lng/41Lng Both Qbs played very much alike, CPU needing a bit better %, besides that they were like mirror images. I really am impressed w/ the CPU decision making in the pass game! Short, Med., Lng, throwaways. Wow! I was playing myself, almost, but not quite. Lil better rush avg. 3.9/3.0- still need a lil CPU boost(apx. 4-6%) Penalties: 1/2(1st half) Both sides forced a fumble, and this was also in similar fashion. S. Moss running over the Middle on both teams and both Laron Landrys SMACK! FUMBLE. The CPU got me first and then, bam I was able to get a bead on Moss running over the middle. It was almost back to back. 2 forced fumbles 1 on each side and all within 3 plays lol. Sacks user/cpu 1-2 Also each side w/ an int! Sweet! This was the set used 50/50 27/39 27/39 50/50 32/44 27/39 27/39 28/39 27/39 27/39 27/39 27/39 I really think just a minor boost to Rblk 3-4% for CPU and 3-6% increase to CPU wr cth, and I bet you will have a really great base sliders(BTW these are considered Allpro bc I can tell by kickmeter speed.) I'm going to test abt 3 more and see the results Oh yeah, score @ halftime in this gm. User/CPU-9-7, I got a last sec. Fg to get the lead bck. 3fgs. I'm sure it would open up a bit more in 2nd half but I don't want fatigue to skew the slider testing. Good stuff!

Looking at your slider groups you are close to being right in between all-pro and all-madden. AP = 150, AM = 75 ... so exactly in between would be 112.5 (total points for each group of three sliders).

Looks like your lowest set is HUM run defense which is closer to AM at 27, 27, 27 = 81 pts, but the other groups are closer to AP at 27, 27, 50 = 104 pts, 27, 32, 50 = 109 pts.

Do you feel challenged by the CPU offense? When I had my sliders set to exactly between AP and AM, that was my problem. I could stop the CPU offense without even controlling a player on defense. I was simply calling the plays and I still shut down the CPU?