User Pass
Home Sign Up Contact Log In
Forum > Suggestions > Points given up instead of Points scored as a tie-breaker
Page:
 
Link
 
If we use points scored as a tiebreaker, it measures a team's ability to run up the score on overmatched teams. The difference between a 150-0 victory and a 200-0 victory has a major effect. The difference between a 30-20 victory and a 40-10 victory has a much smaller effect.

If we do points given up as a tiebreaker, it measures a team's ability to shut down good opponents. Overmatched opponents will all score zero points - there won't be a huge discrepancy on how the top teams beat them in terms of points given up. The difference in this category will come from the competitive games, not from the blowouts.

The tiebreaker should be measured by how you perform against good opponents, not based off of how badly you can blowout bad opponents.
 
Ratphlegm
offline
Link
 
It should be score differential, imo, number of points scored more than number of points given up. Some teams are built to out-score opponents and hope their defense can get a stop or three, others are built to score just enough points and let their stalwart defense shut the opposition down ... score differential makes both these options relatively equal as far as tie breakers are concerned.

I also agree with those who have said only games against the top 8 or 10 teams in the conference should be factored into tiebreaker numbers. That way, you take the unbalanced non-conference game, and the least-competetive teams out of the calculation.
 
Guppy, Inc
offline
Link
 
Originally posted by Ratphlegm
It should be score differential, imo, number of points scored more than number of points given up. Some teams are built to out-score opponents and hope their defense can get a stop or three, others are built to score just enough points and let their stalwart defense shut the opposition down ... score differential makes both these options relatively equal as far as tie breakers are concerned.

I also agree with those who have said only games against the top 8 or 10 teams in the conference should be factored into tiebreaker numbers. That way, you take the unbalanced non-conference game, and the least-competetive teams out of the calculation.


yep, point differential against the playoff teams would be the best way. But that goes against the GLB philosophy of being stats driven
 
jmog
offline
Link
 
I like the points differential vs only the top 10 or so teams as the tie breaker.
 
Jed
offline
Link
 
Originally posted by Ratphlegm
It should be score differential, imo, number of points scored more than number of points given up.


This
 
BiggerBlue
offline
Link
 

Nope. Still wouldn't fix the original argument that it's unfair for teams jacking up scores against gutted teams.

200-0 - Differential 200
150-0 - Differential 150

What's "fix" from using pure Total Points Scored?

If another team happens to be unlucky and play a very balance schedule:

30-21
14-10
28-20
19-9
etc.

The differential is going to be very small. All the other tie-breaker team against you has to do is play a couple of gutted games and that would defeat the whole purpose of points differential.

The only true way to fix this is to find something unique to GLB that is within a fair range for all playoff bound teams across the board, instead of trying to morph NFL into GLB. Some things just can't be move over as easily.

Or, keep the current setting if no real improvement can be think of, at least at this moment I couldn't.
 
Ratphlegm
offline
Link
 
Originally posted by BiggerBlue

Nope. Still wouldn't fix the original argument that it's unfair for teams jacking up scores against gutted teams.

200-0 - Differential 200
150-0 - Differential 150

What's "fix" from using pure Total Points Scored?

If another team happens to be unlucky and play a very balance schedule:

30-21
14-10
28-20
19-9
etc.

The differential is going to be very small. All the other tie-breaker team against you has to do is play a couple of gutted games and that would defeat the whole purpose of points differential.


I guess you're arguing about teams from different conferences who would play completely different opponents, because teams from the same conference play 14 of the same opponents, each other, and one random inter-conference game. I already mentioned that the inter-conference game shouldn't be calculated in the tiebreakers... and that the least competetive teams should also be removed from the calculations.

There's really nothing you can do about deciding who gets homefield in a league championship game if the two teams have identical records and didn't luck into playing each other in their inter-conference game. It probably should be a "neutral" site and have the ticket sales and concessions split evenly, regardless of the records of the participating championship teams.
 
lump7502
offline
Link
 
pts diff with a 21 point cap
 
Link
 
why not take sum of rank in points scored and points allowed?
 
NiborRis
offline
Link
 
It should be points allowed, but only with Cruise Control disabled as that skews all the results.

The main reason points allowed should be more important is that there are lots of game situations where you are not trying to maximize the points scored - rushing a lot with a decent lead, running out the clock on the last drive, whatever, good football tactics are sometimes NOT to score. Much more often you are looking to minimize your opponents' scoring. A few of the corner cases where you would give the opponent points on purpose aren't even in the game (no intentional safeties for starters) so it would be a much better tiebreaker.
 
bmedders
offline
Link
 
+1 to the OP. Even the point differential means it's about how badly you can pounds crappy teams. It especially gets thrown off if team a plays a slow build team and team b plays a top 4 team in the inter-conference game.
 
EagleOtto
offline
Link
 
Originally posted by lump7502
pts diff with a 21 point cap


I like the cap, but it should be higher, 50 would be a good round number, lets you run the score up, but not to unrealistic numbers
 
Link
 
Originally posted by Guppy, Inc
Originally posted by Ratphlegm

It should be score differential, imo, number of points scored more than number of points given up. Some teams are built to out-score opponents and hope their defense can get a stop or three, others are built to score just enough points and let their stalwart defense shut the opposition down ... score differential makes both these options relatively equal as far as tie breakers are concerned.

I also agree with those who have said only games against the top 8 or 10 teams in the conference should be factored into tiebreaker numbers. That way, you take the unbalanced non-conference game, and the least-competetive teams out of the calculation.


yep, point differential against the playoff teams would be the best way. But that goes against the GLB philosophy of being stats driven


Point differential has the exact same problem as points scored. Blowouts against gutted or CPU teams skew the stats way too much.

Counting only games against playoff teams, whether you use points scored, point differential, or points given up is also a good solution. Games against gutted teams should not be the games determining the tiebreakers, it should be games against worthy opponents. It's a reasonable assumption (which will occasionally be wrong) that playoff teams are worthy opponents.
 
didymus
offline
Link
 
points should not be a tie breaker.

I've seriously considered setting my AI to kneel if we are ahead by too much, to just nerf the score on my own.
 
mikey79
offline
Link
 
+1 for Point differential against playoff teams.
 
Page:
 


You are not logged in. Please log in if you want to post a reply.