Nacho Simulation Football League (Season 26 - Taco Bowl XXVI Posted!)

ravens5520 · September 1, 2023

5 hours ago, TheKillerNacho said:

You had like 5 full days...

Work/personal not your deadline

TL-TwoWinsAway · September 1, 2023

Speaking of "MVP formula" over owner vote for the All-Star Game...

There are instances where it seems that "MVP formula" really falls short in identifying the better player. If we were able to - somehow - tweak the formula to account for certain things, using it to decide the ASG would be fantastic.

Examples:

5. Jahmyr Gibbs (ANC): 306 carries for 1,194 yards (3.90 YPC, 31 LNG), 5 touchdowns, 3 fumbles. 43 receptions for 481 yards (11.19 YPR, 31 LNG), 4 touchdowns, 1 drop, 1 fumble.
7. Josh Jacobs (NZ): 269 carries for 1,220 yards (4.54 YPC, 23 LNG), 10 touchdowns, 1 fumble. 27 receptions for 330 yards (12.22 YPR, 23 LNG), 1 touchdown, 3 drops, 2 fumbles.
9. Nick Chubb (RJ): 259 carries for 1,241 yards (4.79 YPC, 39 LNG), 8 touchdowns, 1 fumble. 18 receptions for 222 yards (12.33 YPR, 30 LNG), 2 touchdowns, 4 drops.

I don't think anyone would care to argue that Gibbs was better than both Jacobs and Chubb in S26.

1. Joe Mixon (HAM): 158 carries for 597 yards (3.78 YPC, 21 LNG), 3 touchdowns. 24 receptions for 246 yards (10.25 YPR, 26 LNG), 1 touchdown, 5 drops. 5 pancakes.
2. Jakob Johnson (PHO): 12 carries for 68 yards (5.67 YPC, 14 LNG), 1 touchdown. 11 receptions for 152 yards (13.82 YPR, 37 LNG), 3 touchdowns. 24 pancakes.

A good example here of how easy it would be for certain owners to manipulate the voting. (Of course, I'm not saying that it happened here.) Mixon saw RB1/RB2 carries throughout the season but would've gotten the nod for starting FB. I'd much rather have Johnson in that game, who was both more efficient as a runner and receiver and had tons of pancakes.

5. Ja'Marr Chase (HAM): 87 receptions for 1,077 yards (12.38 YPR, 58 LNG), 10 touchdowns, 6 drops.
6. Justin Jefferson (ANC): 85 receptions for 1,237 yards (14.55 YPR, 37 LNG), 7 touchdowns, 1 drop, 2 fumbles.

Is there some way to deduct for negative plays? The fact that Jefferson is 6th on the list is wild, considering his ridiculous YPR and how infrequently he drops the ball. Comparing him to Chase, I think Jefferson should be ranked above each and every time.

38. Calijah Kancey (WAT): 47 tackles (11 stuffs), 3 coverages, 3 sacks, 2 forced fumbles, 3 recoveries.
39. Dexter Lawrence (LM): 68 tackles (19 stuffs), 1 coverage, 3 sacks, 1 forced fumble.

Not to harp on this one, but I can't see Kancey being above Lawrence here. The disparity between the tackles and stuffs is massive. (I thought Lawrence should've been much higher than 39th. Perhaps more of an emphasis on stuffs?)

Just some random thoughts.

TheKillerNacho · September 1, 2023

Quote

Speaking of "MVP formula" over owner vote for the All-Star Game...

There are instances where it seems that "MVP formula" really falls short in identifying the better player. If we were able to - somehow - tweak the formula to account for certain things, using it to decide the ASG would be fantastic.

Well, do bear in mind its meant to find Game MVPs and sometimes that can be quite different then identifying better seasonal stats. While it is intended to be used for both, its often a balancing act between the two. I do thank you for your time compiling these concerns, though (and perhaps some things could be tweaked)

44 minutes ago, TL-TwoWinsAway said:

Examples:

5. Jahmyr Gibbs (ANC): 306 carries for 1,194 yards (3.90 YPC, 31 LNG), 5 touchdowns, 3 fumbles. 43 receptions for 481 yards (11.19 YPR, 31 LNG), 4 touchdowns, 1 drop, 1 fumble.
7. Josh Jacobs (NZ): 269 carries for 1,220 yards (4.54 YPC, 23 LNG), 10 touchdowns, 1 fumble. 27 receptions for 330 yards (12.22 YPR, 23 LNG), 1 touchdown, 3 drops, 2 fumbles.
9. Nick Chubb (RJ): 259 carries for 1,241 yards (4.79 YPC, 39 LNG), 8 touchdowns, 1 fumble. 18 receptions for 222 yards (12.33 YPR, 30 LNG), 2 touchdowns, 4 drops.

I don't think anyone would care to argue that Gibbs was better than both Jacobs and Chubb in S26.

The difference in receiving stats is significant (Gibbs has almost as many receptions as Jacobs and Chubb combined and only 1 drop compared to 3+), so it makes some sense. And due to the receiving yardage, Gibbs has significantly more bulk yardage (1,675) than Jacobs (1,550) or Chubb (1,463). Bulk stats are the only thing the formula cares about as its hard to assign a value to efficiency stats like YPC that would work for both a single game and a whole season. This was one of the primary reason I previously favored owner votes for the All-Star Bowl - the Game MVP formula is indeed poor at evaluating efficiency.

Quote

1. Joe Mixon (HAM): 158 carries for 597 yards (3.78 YPC, 21 LNG), 3 touchdowns. 24 receptions for 246 yards (10.25 YPR, 26 LNG), 1 touchdown, 5 drops. 5 pancakes.
2. Jakob Johnson (PHO): 12 carries for 68 yards (5.67 YPC, 14 LNG), 1 touchdown. 11 receptions for 152 yards (13.82 YPR, 37 LNG), 3 touchdowns. 24 pancakes.

A good example here of how easy it would be for certain owners to manipulate the voting. (Of course, I'm not saying that it happened here.) Mixon saw RB1/RB2 carries throughout the season but would've gotten the nod for starting FB. I'd much rather have Johnson in that game, who was both more efficient as a runner and receiver and had tons of pancakes.

This is the type of thing that I'll account for when/if the Game MVP formula takes over for All-Star Bowl rosters (disqualifying players who spent a large portion of their time at another position that typically accumulates more stats). In I had done the All-Star Bowl rosters with the formula this time, I would've simply removed Mixon from consideration or thrown him in with the RBs.

Quote

5. Ja'Marr Chase (HAM): 87 receptions for 1,077 yards (12.38 YPR, 58 LNG), 10 touchdowns, 6 drops.
6. Justin Jefferson (ANC): 85 receptions for 1,237 yards (14.55 YPR, 37 LNG), 7 touchdowns, 1 drop, 2 fumbles.

Is there some way to deduct for negative plays? The fact that Jefferson is 6th on the list is wild, considering his ridiculous YPR and how infrequently he drops the ball. Comparing him to Chase, I think Jefferson should be ranked above each and every time.

It already does deduct for negative plays like drops and fumbles, although keep in mind the main purpose is for single games so bulk stats for a whole season often have these negatives midigated. Also notably, the formula's negative weight for fumbles (-2.1) is far greater than its negative weight for drops (-0.4) so Jefferson (-4.6) is actually being punished more than Chase (-2.4) for his negative plays. Perhaps this is something that could indeed be adjusted, but I always felt potential turnovers should be punished more than simply a dropped ball.

Quote

38. Calijah Kancey (WAT): 47 tackles (11 stuffs), 3 coverages, 3 sacks, 2 forced fumbles, 3 recoveries.
39. Dexter Lawrence (LM): 68 tackles (19 stuffs), 1 coverage, 3 sacks, 1 forced fumble.

Not to harp on this one, but I can't see Kancey being above Lawrence here. The disparity between the tackles and stuffs is massive. (I thought Lawrence should've been much higher than 39th. Perhaps more of an emphasis on stuffs?)

Turnovers are just valued really highly, to be honest. Also since coverages for DL are mainly used for tipped balls, I give DL a bit of a boost for them. But it does seem weird to me too, i'd have to math it out

If it helps for discussion purposes, I can just link the Game MVP formula itself (I don't think this is really super sensitive information). Although please excuse the horrible spaghetti coding and poor choice for some of the internal variable stat names (a great many of them are named differently internally than what is spit out for stat output), keeping in mind I started this project when I was a very very new programmer.

public void calculateRating() {
	double puntNum = 0.0;
	if (punts > 0)
	{
		puntNum = 45.0;
		if (punts < 3)
			puntNum += (3 - punts) * 5.0;
	}
	double returnNum = 0.0;
	if (returns > 0)
		returnNum = 12.5;
	double pancakeNum = 1.0;
	if (listedPos == SeasonDraft.FB)
		pancakeNum = 1.25;
	if (listedPos == SeasonDraft.OT || listedPos == SeasonDraft.OG || listedPos == SeasonDraft.C)
		pancakeNum = 1.5;

	//tfls were 0.75 and 0.25 extra for DL (and 0.115 for tackles extra) prior to stuff change. tflAllowed was 0.45
	rating = (1.0 * (yardsPassing / 44.25)) //previosly 43.375
		+ (2.25 * tdsPassing)
		+ (4.0 * interceptions) + (0.9 * fgMade) + (0.05 * xpMade) + (2.75 * sacks)
		+ (3.0 * forcedFumble) + (2.5 * fumRec) + (4.5 * safety)
		+ (0.25 * tackles) + (0.55 * tfls) + (0.385 * coverages) + (5.0 * tdDefense)
		+ (3.0 * returnTD) + (0.3 * touchbacks) + (0.6 * in20) + (pancakeNum * pancakes)
		+ (0.25 * ((double)returnYards / ((double)returns + 1) - returnNum)) + (2.5 * tdsCatching)
		+ (0.35 * (averagePuntDouble() - puntNum)) - (2.75 * (interceptionsThrown))
		- (2.15 * (fumbles)) - (2.1 * (fumbles2)) - (1.5 * (fgTrys - fgMade))
		- (2.0 * (xpTrys - xpMade)) - (0.9 * sacksAllowed) - (0.3 * tflAllowed)
		- (0.00325 * (attempts - completions));

	if (listedPos != SeasonDraft.QB) {
		rating = rating + (1.0 * (yardsRushing / 20.325))
				+ (2.5 * tdsRushing);
	} else {
		rating = rating + (1.0 * (yardsRushing / 50.0))
				+ (2.0 * tdsRushing);
	}

	if (listedPos == SeasonDraft.DE || listedPos == SeasonDraft.DT) {
		rating += (0.777 * coverages) + (0.07625 * tfls) + (0.115 * tackles);
	}

	if (listedPos != SeasonDraft.RB) {
		double dropWeight = 0.4;
		double catchWeight = 0.32275;
		if (listedPos == SeasonDraft.FB)
		{
			catchWeight = 0.14863;
			dropWeight = 0.05;
		}
		if (listedPos == SeasonDraft.TE)
		{
			dropWeight = 0.2;
		}
		rating = rating + (catchWeight * catches) + (0.2255 * tdsCatching)  + (1.0 * (yardsCatching / 20.0) - (dropWeight * drops));
	} else {
		rating = rating + (0.14863 * catches) + (1.0 * (yardsCatching / 20.325) - (0.1 * drops));
	}

	rating *= 10.0;
	if (name.equals("Null Player") && team.name.equals("Null Team"))
		rating = -666;
}

(also to answer a potential question before it arises - returner stats are removed for the season stat ordering other than Returner obviously, to avoid giving Returners an advantage there)

TL-TwoWinsAway · September 1, 2023

6 minutes ago, TheKillerNacho said:
Well, do bear in mind its meant to find Game MVPs and sometimes that can be quite different then identifying better seasonal stats. I do thank you for your time compiling these concerns, though (and perhaps some things could be tweaked)

The difference in receiving stats is significant (Gibbs has almost as many receptions as Jacobs and Chubb combined and only 1 drop compared to 3+), so it makes some sense. Gibbs has more bulk yardage (1,675) than Jacobs (1,550) or Chubb (1,463) and fewer drops. Bulk stats are the only thing the formula cares about as its hard to assign a value to efficiency stats like YPC that would work for both a single game and a whole season. This was one of the primary reason I previously favored owner votes for the All-Star Bowl - the Game MVP formula is indeed poor at evaluating efficiency.

This is the type of thing that I'll account for when/if the Game MVP formula takes over for All-Star Bowl rosters (disqualifying players who spent a large portion of their time at another position that typically accumulates more stats). In I had done the All-Star Bowl rosters with the formula this time, I would've simply removed Mixon from consideration or thrown him in with the RBs.

It already does deduct for negative plays like drops and fumbles, although keep in mind the main purpose is for single games so bulk stats for a whole season often have these negatives midigated. Also notably, the formula's negative weight for fumbles (-2.15) is far greater than its negative weight for (-0.4) so Jefferson (-4.7) is actually being punished more than Chase (-2.4) for his negative plays. Perhaps this is something that could indeed be adjusted, but I always felt potential turnovers should be punished more than simply a dropped ball.

Turnovers are just valued really highly, to be honest. Also since coverages for DL are mainly used for tipped balls, I give DL a bit of a boost for them. But it does seem weird to me too, i'd have to math it out

If it helps for discussion purposes, I can just link the Game MVP formula itself (I don't think this is really super sensitive information). Although please excuse the horrible spaghetti coding and poor choice for some of the internal variable stat names (keeping in mind I started this project when I was a very very new programmer).
public void calculateRating() {
		double puntNum = 0.0;
		if (punts > 0)
		{
			puntNum = 45.0;
			if (punts < 3)
				puntNum += (3 - punts) * 5.0;
		}
		double returnNum = 0.0;
		if (returns > 0)
			returnNum = 12.5;
		double pancakeNum = 1.0;
		if (listedPos == SeasonDraft.FB)
			pancakeNum = 1.25;
		if (listedPos == SeasonDraft.OT || listedPos == SeasonDraft.OG || listedPos == SeasonDraft.C)
			pancakeNum = 1.5;
		
		//tfls were 0.75 and 0.25 extra for DL (and 0.115 for tackles extra) prior to stuff change. tflAllowed was 0.45
		rating = (1.0 * (yardsPassing / 44.25)) //previosly 43.375
				+ (2.25 * tdsPassing)
				+ (4.0 * interceptions) + (0.9 * fgMade) + (0.05 * xpMade) + (2.75 * sacks)
				+ (3.0 * forcedFumble) + (2.5 * fumRec) + (4.5 * safety)
				+ (0.25 * tackles) + (0.55 * tfls) + (0.385 * coverages) + (5.0 * tdDefense)
				+ (3.0 * returnTD) + (0.3 * touchbacks) + (0.6 * in20) + (pancakeNum * pancakes)
				+ (0.25 * ((double)returnYards / ((double)returns + 1) - returnNum)) + (2.5 * tdsCatching)
				+ (0.35 * (averagePuntDouble() - puntNum)) - (2.75 * (interceptionsThrown))
				- (2.15 * (fumbles)) - (2.1 * (fumbles2)) - (1.5 * (fgTrys - fgMade))
				- (2.0 * (xpTrys - xpMade)) - (0.9 * sacksAllowed) - (0.3 * tflAllowed)
				- (0.00325 * (attempts - completions));

		if (listedPos != SeasonDraft.QB) {
			rating = rating + (1.0 * (yardsRushing / 20.325))
					+ (2.5 * tdsRushing);
		} else {
			rating = rating + (1.0 * (yardsRushing / 50.0))
					+ (2.0 * tdsRushing);
		}
		
		if (listedPos == SeasonDraft.DE || listedPos == SeasonDraft.DT) {
			rating += (0.777 * coverages) + (0.07625 * tfls) + (0.115 * tackles);
		}

		if (listedPos != SeasonDraft.RB) {
			double dropWeight = 0.4;
			double catchWeight = 0.32275;
			if (listedPos == SeasonDraft.FB)
			{
				catchWeight = 0.14863;
				dropWeight = 0.05;
			}
			if (listedPos == SeasonDraft.TE)
			{
				dropWeight = 0.2;
			}
			rating = rating + (catchWeight * catches) + (0.2255 * tdsCatching)  + (1.0 * (yardsCatching / 20.0) - (dropWeight * drops));
		} else {
			rating = rating + (0.14863 * catches) + (1.0 * (yardsCatching / 20.325) - (0.1 * drops));
		}

		rating *= 10.0;
		if (name.equals("Null Player") && team.name.equals("Null Team"))
			rating = -666;

}

//Added to the above for Game MVPs and used exclusively for seasonal Returner ordering, but not used for other seasonal ordering to avoid giving an advantage to returners in those lists.
public void calculateKRRating() {
		double returnNum = 0.0;
		if (returns > 0)
			returnNum = 12.5;
		KRrating = (3.0 * returnTD)
				+ (0.25 * ((double)returnYards / ((double)returns + 1) - returnNum));
		//System.out.println(name + " " + KRrating);
}

Makes perfect sense. Thanks for the info!

Edit: and maybe, as you suggested this year, submitting a "MVP formula" ballot along with any ballots submitted by owners would help balance things a bit.

Edited September 1, 2023 by TL-TwoWinsAway

LAOJoe · September 1, 2023

15 minutes ago, TheKillerNacho said:

but I always felt potential turnovers should be punished more than simply a dropped ball.

Getting and turning over the ball was a big part of my voting and why I had some players higher where others wouldn't.

Maybe voting should be mandatory without an advance reason given. Maybe everyone starts with 19 tags instead of 20 starting next season and gets a tag for at least voting for their conference or giving a valid excuse. I'd rather that than a formula.

TL-TwoWinsAway · September 1, 2023

20 minutes ago, TheKillerNacho said:

Turnovers are just valued really highly, to be honest. Also since coverages for DL are mainly used for tipped balls, I give DL a bit of a boost for them. But it does seem weird to me too, i'd have to math it out

I suck at numbers, but I think I did the maths here:

Kancey
47 tackles: 11.75
11 stuffs: 6.05
3 coverages: 1.155
3 sacks: 8.25
2 forced fumbles: 6
3 recoveries: 7.5
----------------------
40.705

Lawrence
68 tackles: 17
19 stuffs: 10.45
1 coverage: 0.385
3 sacks: 8.25
1 forced fumble: 3.0
------------------
39.085

Just edged him out. The recoveries carried Kancey here. I feel like stuffs should be weighed far higher for DL (even as high as 1.0 from 0.55, making them four times as valuable as just a tackle), but that might be too much. I don't know.

TheKillerNacho · September 1, 2023

15 minutes ago, LAOJoe said:

Getting and turning over the ball was a big part of my voting and why I had some players higher where others wouldn't.

Maybe voting should be mandatory without an advance reason given. Maybe everyone starts with 19 tags instead of 20 starting next season and gets a tag for at least voting for their conference or giving a valid excuse. I'd rather that than a formula.

thank you for submitting a thoughtful ballot, I do appreciate it.

i'll mull the options over for next season, but I'm unsure I like forcibly punishing those who dont submit a ballot. And i think it would just lead to owners who would otherwise not feel like submitting one to just submit their whole roster or copypaste someone else's or something.

perhaps a combination of using the formula & owner voting is best. idk

TL-TwoWinsAway · September 1, 2023

Just now, TheKillerNacho said:

i'll mull the options over for next season, but I'm unsure I like forcibly punishing those who dont submit a ballot. And i think it would just lead to owners who would otherwise not feel like submitting one to just submit their whole roster or copypaste someone else's or something.

perhaps a combination of using the formula & owner voting is best. idk

This, exactly. I'd rather no ballot over someone just submitting their own team.

And I think submitting an MVP formula ballot (even with inflated scoring) along with the owner ballots would work.

TheKillerNacho · September 1, 2023

3 minutes ago, TL-TwoWinsAway said:

I suck at numbers, but I think I did the maths here:

Kancey
47 tackles: 11.75
11 stuffs: 6.05
3 coverages: 1.155
3 sacks: 8.25
2 forced fumbles: 6
3 recoveries: 7.5
----------------------
40.705

Lawrence
68 tackles: 17
19 stuffs: 10.45
1 coverage: 0.385
3 sacks: 8.25
1 forced fumble: 3.0
------------------
39.085

Just edged him out. The recoveries carried Kancey here. I feel like stuffs should be weighed far higher for DL (even as high as 1.0 from 0.55, making them four times as valuable as just a tackle), but that might be too much. I don't know.

did you miss this?

if (listedPos == SeasonDraft.DE || listedPos == SeasonDraft.DT) {
		rating += (0.777 * coverages) + (0.07625 * tfls) + (0.115 * tackles);
	}

so yeah the coverages actually are even more critical for Kancey, here. But the TFLs are worth slightly more too

TL-TwoWinsAway · September 1, 2023

Just now, TheKillerNacho said:

did you miss this?

if (listedPos == SeasonDraft.DE || listedPos == SeasonDraft.DT) {
		rating += (0.777 * coverages) + (0.07625 * tfls) + (0.115 * tackles);
	}

Oh. Yep. Of course I did. I'll re-math.

TheKillerNacho · September 1, 2023

The appetizer to the Taco Bowl is here, the All-Star Bowl XXIII! Interestingly, the conferences each have had 11 victories in the game all-time, so which will break the tie? And who will get the measly yet coveted Keeper Tag prize to spend on the Game's MVP next season?

All-Star Bowl XXIII: https://youtu.be/J8ZZXhp9Ork
GameDoc: https://drive.google.com/file/d/1fW3LVZTzpOJQfQqETgDqbOc6C4XTuL74/view?usp=sharing

As for the "bizarro game" I mentioned, I'm busier than I thought I'd be today and no one seemed really all that interested so I didn't really bother. I can still do it later, maybe.

TL-TwoWinsAway · September 1, 2023

Spoiler

Kingram wins despite Cousins' best efforts. Well played, good sir.

Murphy-Bunting over here getting the last laugh:

Sean Murphy-Bunting: 1 tackle, 6 coverages, 1 interception.

Malfatron · September 1, 2023

6 minutes ago, TL-TwoWinsAway said:
Hide contents
Kingram wins despite Cousins' best efforts. Well played, good sir.

Murphy-Bunting over here getting the last laugh:
Sean Murphy-Bunting: 1 tackle, 6 coverages, 1 interception.

Spoiler

At least we didnt lose because of Kirk. He did pretty good!

LAOJoe · September 1, 2023

5 minutes ago, TL-TwoWinsAway said:
Hide contents
Kingram wins despite Cousins' best efforts. Well played, good sir.

Murphy-Bunting over here getting the last laugh:
Sean Murphy-Bunting: 1 tackle, 6 coverages, 1 interception.

Spoiler

Should have been the real MVP over Henry even with 100 yards.

Wish Cheesy went for it on 4th down. Kept getting short yardage on 3rd down runs and being short. Also only 1 pancake by the FB. The defensive line for Salsa was too stacked.

Malfatron · September 1, 2023

1 minute ago, LAOJoe said:

Hide contents

Should have been the real MVP over Henry even with 100 yards.

Wish Cheesy went for it on 4th down. Kept getting short yardage on 3rd down runs and being short. Also only 1 pancake by the FB. The defensive line for Salsa was too stacked.

Spoiler

Weird that Barkley got MVP even with worse stats. Edit: ah its probably the 2 catches

Edited September 1, 2023 by Malfatron

Nacho Simulation Football League (Season 26 - Taco Bowl XXVI Posted!)

Recommended Posts

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites