http://www.operationsports.com/forum...tatistics.html
Quote:
I ran 40 yard dash in practice mode with offense only with >90 speed and acceleration vs 80 speed and acceleration. I tested this from a standstill with a stopwatch. Real world values were gathered from the NFL combines published numbers for all LB and RB. I used LB and RB because this coverage in the flats is the main issue. I tested for Default settings and normal speed, slow speed with default threshold and 0 and 100 threshold.
See the table below:
Real
world
Default Slow 100
thresh
0
thresh
HB 4.53 4.53 4.88 4.48 4.34
LB 4.69 4.72 5.11 4.68 4.9
The results show that the default sliders are accurate for RB 40 times but not for LB though actually not very far off. Slow speed setting markedly leads to unrealistic values for both thoug it maintains the 0.2 sec difference. I do not recommend slow if you want real world values, but it maintains the balance and should not effect gameplay. The 100 threshold slider did not actually narrow the gap anymore but did improve the LB speed (I noticed acceleration was having the biggest impact; I also suspec the HB values may not be accurate, median was still 4.55). 0 threshold had a massive impact on decreasing HB time and increasing LB time widening the gap to 0.6 sec difference. Increasing the threshold value towards 100 should result in more realistic 40 times and hence linebacker coverage in the flats. Tonight I will perform testing at 60/70/80/90 to find best value with 20 measurements each.
Summary: Speed should be normal for realistic running times, Threshold slider at 50 is actually
Quote:
Here are the finished results. I finally decided to perform the test from the 35 yard line since there is no block and jump like in real 40 time and this gave close to reealistic HB and LB times. I used a LB with 80 stats for accel and speed and HB with 90 accel and 90 speed. 40 time mostly reflects acceleration so unsure of this impact on deep routes but I feel they need to adjust acceleration for this game as this stat is a bit overpowered currently.
Real NFL statistics:
average time LB: 4.7 sec
average time HB: 4.5 sec
average time WR: 4.5-4.6 sec
average difference LB and HB: 0.2 sec
NCAA football stastics:
Slider HB LB Diff
\50 4.341 4.760 -0.419
60 4.332 4.639 -0.307
70 4.326 4.614 -0.288
80 4.345 4.577 -0.232
90 4.377 4.584 -0.207
100 4.356 4.524 -0.168
You can easily see that this followed a very nice progression of about 0.4-0.5 decrease int he gap per 10 change in the slider. My 40 times are not realiable since they are estimates of times so the most important factor is the difference between the HB and LB. You can see that the NFL combine statistics would require a slider setting of 90 to be accurate. However, it is well know that college defenses are slower than NFL which is a main difference therefore the likely gap for NCAA should be around 0.250. I am going to set my sliders at 75 which will likley give a gap of around 0.250 and if my values for HB and LB were off by 0.15 (diff between real HB and my stats) the linebacker speed would be around 4.7 (4.60 + 0.15).
Summary: Slider settings between 70 and 80 should give real world 40 times to players and allow LB to cover both WR, TE, and HB. Hopefully this will improve coverage on crossing routes and flats.
Next: Will subjectively examine the impact of 0, 50, 75, 100 on flats and crossing routes against man and zone coverage.
Quote:
After subjective testing I have found the following to be true:
1. Coverage of crossing and drag routes is poor as these are almost 100% effective against man coverage.
2. Zone coverage is relatively uneffected by this though I noted less gains with LB closing more quickly.
3. This is insanely hard. LB can remain in fairly close coverage of even wide recievers downfield so big gains are remarkably reduced. Downfield passing seems extremely difficult. This can be fixed by reducing the pass coverage slider and a setting of 40 seemed to open up passing a bit more. Unfortunately this is for human. The CPU seemed to not have much difficulty finding open recievers still (on short routes). I have ultimately decided to go with a slider setting of 70.
Quote:
This is on AA. The developers have stated for several years that the game is not intended to be sim on Heisman as the CPU has unrealistic completeoin % and decision making etc... on this level. They have previously stated that AA is the level they shoot for making realistic.