NOTICE: This league is using the BLEEDING EDGE game engine. For more information, click here.

The new user interface is in preview!

Want to check it out? Click here! (If you don't like it, you can still switch back)

League Forums

Main - Beta Chat

Re: 0.4.4 Observations

By setherick
12/08/2018 7:24 am
The commit thread is getting cluttered, so I started a new one for general observations about 0.4.4.

I'm still seeing a lot of plays like this. Where is the QB looking? And why does he not look at the TE here? In one second, this play is going to end in a Safety. QBs still need to make decisions on medium routes faster. If a player has separation, why are they waiting to throw the ball?

In all fairness, the QB on this play has terrible release.



https://private75.myfootballnow.com/watch/3698#694770

Here's another play from the same game that illustrates something else that I've seen. Players are unpredictable in how they run drag routes. Most of the time they cross across the face of the QB, but some times they take a hard turn in the other direction. Look at the two screens below. In the first one, you expect the WR to keep running across the field (white arrow), but the red arrow is actually where he goes. It's like he gets confused when the RB crosses his path and starts running the RBs route.



https://private75.myfootballnow.com/watch/3698#694782
Last edited at 12/08/2018 7:32 am

Re: 0.4.4 Observations

By raymattison21
12/08/2018 7:48 am
Bump and man all slow the receiver down too much and Oline blocking is bad at times. That's messing with timing. QBs making bad reads is up for debate . If they threw some quicker underneath routes on long and medium passes I would feel better. Cause I can hit backs and TEs out of big and base sets, but spread sets with no backs going out seem to be a sack instead of an easy completion underneath . Also, so many drops on long passes really distorts things from a complete perspective . Playing deep might be only reserved for lack of talent .

Re: 0.4.4 Observations

By Warthog
12/08/2018 8:51 am
Something I just realized is that completion percentages are through the roof. My QB in Rivals last night started the game 29-29. Then I looked at all the QBs in the league and like seven or eight are completing 80% of their passes. That is WAY too high.

Re: 0.4.4 Observations

By setherick
12/08/2018 8:53 am
Warthog wrote:
Something I just realized is that completion percentages are through the roof. My QB in Rivals last night started the game 29-29. Then I looked at all the QBs in the league and like seven or eight are completing 80% of their passes. That is WAY too high.


I think you'd see the same thing in the NFL if QBs played the same way they do in MFN. Percentages are high and yard per attempt are low because QBs check down to open WRs and RBs. It's a trade off.

If RBs were staying in to block more, you'd see a higher yard per attempt and lower completion percentage.

Re: 0.4.4 Observations

By setherick
12/08/2018 9:52 am
Actually, I have a better idea for how to adjust completion percentages.

Right now, QBs determine to throw based on a probability matrix, but I'm not exactly sure how that works. Here's how I think it should work.

1) The probability of a particular throw being completed is calculated. This value is hidden from the QB.

2) The QB determines the probability inversely to his INT. So an INT of 100 will see the real value determined by the probability calculation. But the lower the INT the more skewed UPWARD the probability will be. That will make lower INT QBs think the throw is better than it is. The trick here would be balancing it out so that a low INT QB didn't throw an INT to the first covered receiver he saw.

3) For low probability throws, especially deep passes, the QB should roll against his Discipline to see if he makes the throw or not. A high Discipline means the QB moves on from the throw, but a low Discipline would make the QB gamble more. This would allow Accurate, Strong Arm, high Intelligence, but low Discipline QBs to be really valuable to teams that wanted to use a deep passing strategy. But would mean that volume passing would require QBs to be Intelligent and Disciplined as well as accurate.
Last edited at 12/08/2018 9:52 am

Re: 0.4.4 Observations

By setherick
12/08/2018 2:18 pm
There is a bug with the 212 Hard Slants. If the QB can't find a receiver, he doesn't do anything, include throw the ball away: https://rivals.myfootballnow.com/watch/117#21071

This happened twice to my opponent. The QB has 81 Release and 64 Scramble, so enough of both to make a fast decision or throw it away.

Re: 0.4.4 Observations

By setherick
12/09/2018 10:27 am
setherick wrote:
Warthog wrote:
Something I just realized is that completion percentages are through the roof. My QB in Rivals last night started the game 29-29. Then I looked at all the QBs in the league and like seven or eight are completing 80% of their passes. That is WAY too high.


I think you'd see the same thing in the NFL if QBs played the same way they do in MFN. Percentages are high and yard per attempt are low because QBs check down to open WRs and RBs. It's a trade off.

If RBs were staying in to block more, you'd see a higher yard per attempt and lower completion percentage.


To my point, look at the NFL leaders right now: https://www.teamrankings.com/nfl/player-stat/completion-percentage

MFN is not that far off.

Also, I think completion percentages are from QBs being overgenerated to make up for other problems in the code. the bell curve for QB should be that 90% should fall into the 50-70 accuracy range.
Last edited at 12/09/2018 10:34 am

Re: 0.4.4 Observations

By raymattison21
12/09/2018 10:40 am
setherick wrote:
Actually, I have a better idea for how to adjust completion percentages.

Right now, QBs determine to throw based on a probability matrix, but I'm not exactly sure how that works. Here's how I think it should work.

1) The probability of a particular throw being completed is calculated. This value is hidden from the QB.

2) The QB determines the probability inversely to his INT. So an INT of 100 will see the real value determined by the probability calculation. But the lower the INT the more skewed UPWARD the probability will be. That will make lower INT QBs think the throw is better than it is. The trick here would be balancing it out so that a low INT QB didn't throw an INT to the first covered receiver he saw.

3) For low probability throws, especially deep passes, the QB should roll against his Discipline to see if he makes the throw or not. A high Discipline means the QB moves on from the throw, but a low Discipline would make the QB gamble more. This would allow Accurate, Strong Arm, high Intelligence, but low Discipline QBs to be really valuable to teams that wanted to use a deep passing strategy. But would mean that volume passing would require QBs to be Intelligent and Disciplined as well as accurate.


I like the idea but think is a bit early too add ratings that cause other effects , cause right now the closest thing is look off. But that puts a defense out of place to make a throw . Not really a gamble at all.

Re: 0.4.4 Observations

By setherick
12/09/2018 10:41 am
Look off already works. It pulls the safety off of the play. The problem is that QBs are gun shy right now on long passes. That's why I'd like to see them roll against Discipline to make downfield throws that aren't perfect.
Last edited at 12/09/2018 10:41 am

Re: 0.4.4 Observations

By raymattison21
12/09/2018 12:46 pm
setherick wrote:
Look off already works. It pulls the safety off of the play. The problem is that QBs are gun shy right now on long passes. That's why I'd like to see them roll against Discipline to make downfield throws that aren't perfect.


And deep wide open guys drop it too often.