Beginning Discussion on a Ball Ladder Fantasy League

[Threevenge]

With the tbd fantasy league up and running, I've heard a few people express interest in running one for ball ladder as well. In order to do so, there's a lot that would need to be discussed and sorted out beforehand...specifically, a possible formula to calculate the value of a player. Which has a whole lot of problems.

The main difficulty with this is ball players generally come in two types: the killers and the scorers.

This usually leads to two different types of players that are both considered "good" : those with a low (sometimes like a 0.5) kill-death ratio but who get something like 1.5 goals per game, and those who rarely score but average around a 1.5 kill-death ratio and who can stop a push on their own. Scoring is obviously the most important part in ball, but without the killers it's near impossible. So, a formula would have to not overemphasize either aspect of the game.

Another issue is kill assists and pass assists. Both have some sort of value, but in comparison to a kill or goal it's a bit unclear.

Anyways, this is just a starting point to see if there's any logical suggestions on a formula and if there's enough interest to run fantasy ball ladder.

[sunshineduck]

problem with this imo (as i explained in mumble) is that any formula that doesn't put shmo or mattman far far ahead of the pack much like donk and rib were simply doesn't work imo. the problem with that is that neither of them has overwhelmingly good stats, which lends to the idea that ball ladder statistics don't necessarily correlate with skill.

i didn't like nade's idea for tbd, but it makes sense for ball if you properly weight it based on rating. the problem with making it an additive multiplier is that the ratings wouldn't be consistent and that as ball ladder inflates the point values of everyone playing would go higher. a way to combat this would be to make it subtract points based on how many points are between the player in question and the rating points required for the #1 spot.

[nobodyhome]

i know that neither kills nor goals scored is a very good indication of how much better shmo and vip are than anyone else. have you looked into correlating maybe ball receptions or ball time carried for this? i know i've looked into this in the past and it seems that better players tend to have a lot of ball receptions.

[lamster]

It's definitely good to think about the best stats to track for fantasy purposes but picking stats that correlate perfectly with impact on the game absolutely is not required. I'd say game impact should be the first criteria, followed very closely by the ease with which a particular stat can be assessed by spectators. Fantasy isn't about picking the best players (that's what ladder is for) it's about gaming the system -- e.g. this quarterback isn't that great but his team sucks so he'll ALWAYS be playing from behind and generating big yardage -- or I hate playing on this guy's team in ball ladder but he always manages to score 3 goals (while letting the opponents walk over us), so I'll pick him up for my fantasy roster. Fantasy is about meta-gaming -- pick some stats that are exciting to watch for and track and just have fun with it

[.sf|hubris]

Originally Posted by Threevenge

This usually leads to two different types of players that are both considered "good" : those with a low (sometimes like a 0.5) kill-death ratio but who get something like 1.5 goals per game, and those who rarely score but average around a 1.5 kill-death ratio and who can stop a push on their own. Scoring is obviously the most important part in ball, but without the killers it's near impossible. So, a formula would have to not overemphasize either aspect of the game.

What if you had offensive and defensive slots to fill on your roster - with the relevant stats being the focus for those players. Similar to a fball fantasy format...

[ryebone]

Another option is to count all the individual stats separately: goals, assists, kills, passes received/completed, etc, and run a rotisserie-style format (go google it if you don't know what that means). Also, I want to re-emphasize lamster's point about Fantasy Ladder being more about the meta-game than actual skill. Fantasy sports is NOT MEANT to be a measure of skill; it's a measure of who can put up the most gaudy stats, and that's it. Ladder rating CAN be one component of the overall formula, but it should not be the main focus.

[nobodyhome]

Originally Posted by ryebone

Fantasy sports is NOT MEANT to be a measure of skill; it's a measure of who can put up the most gaudy stats, and that's it. Ladder rating CAN be one component of the overall formula, but it should not be the main focus.

Agreed. Like I've always said, in fantasy basketball, chris paul and kevin durant are some of the most desired players, while in the actual NBA they are pretty ****ty players overall, and that's ok.

[Smushface]

Originally Posted by ryebone

Another option is to count all the individual stats separately: goals, assists, kills, passes received/completed, etc, and run a rotisserie-style format (go google it if you don't know what that means). Also, I want to re-emphasize lamster's point about Fantasy Ladder being more about the meta-game than actual skill. Fantasy sports is NOT MEANT to be a measure of skill; it's a measure of who can put up the most gaudy stats, and that's it. Ladder rating CAN be one component of the overall formula, but it should not be the main focus.

I'm a big fan of rotisserie chickens.

You can pick up a whole rotisserie chicken from your local grocery store for about $6. I like eating it when it's hot so I get my roommates together, and we have a nice chicken dinner that night. We also make a point not to pick the chicken clean so we throw the bones and leftovers into a big pot with some water, seasoning, and maybe a few veggies (like celery and carrots) and let it simmer for a few hours. Then you filter out all the solids and BOOM! you got yourself a good meal for a couple of people and home made chicken stock.

Like I said, I love the rotisserie style.

[sunshineduck]

Originally Posted by nobodyhome

Agreed. Like I've always said, in fantasy basketball, chris paul and kevin durant are some of the most desired players, while in the actual NBA they are pretty ****ty players overall, and that's ok.

..........

[elixirwithani]

nobo making a hard push to take mikesol's biggest troll crown

[Threevenge]

After reviewing how rotisserie league scoring works, I think that might be an entertaining thing for us to consider. There's several possibilities for statistical categories that we could look at using ladder:

-Goals
-Goal Assists
-Kills
-Kill Assists
-Completed Passes
-Possession Time
-Receptions

-And, for the lols, deaths (I'd be a good pick for this, apparently I'm in the top 10 xD)


I guess the question is how "realistic" do we want this to be in reflecting the quality of the fantasy teams?

[Disilord]

Originally Posted by Threevenge

After reviewing how rotisserie league scoring works, I think that might be an entertaining thing for us to consider. There's several possibilities for statistical categories that we could look at using ladder:

-Goals
-Goal Assists
-Kills
-Kill Assists
-Completed Passes
-Possession Time
-Receptions

-And, for the lols, deaths (I'd be a good pick for this, apparently I'm in the top 10 xD)


I guess the question is how "realistic" do we want this to be in reflecting the quality of the fantasy teams?

I think you should somehow include win percentage for each week.
OR

You should scrap 1 formula for all plain stats and create different formula's for each plain. This however would:
1. Be hard to create the formulas that balance the plains out.
2. Would prevent teams from choosing players who whore the match, such as biplains.

OR

We do the same as the second suggestion but make it so that the managers of the teams have to chose which plain/formula they are going to put for a week.

OR

I'm in the midst of an idea. Be right back

[Aki1024]

To help figure out what multipliers do to what stats, edit this url to what you want to see.

altitudeladder.com/fantasy.php
?mode=ball_6v6
&kills=1
&assists=0.4
&deaths=-0.6
&carrier=1
&xp=0.01
&goals=1
&goalassists=0.5
&receptions=1
&completepasses=1
&recoveries=0

Any values left out default to the current formula (which for ball is all zeros). Values allow decimals, and may be negative or zero.

[Disilord]

Originally Posted by Aki1024

To help figure out what multipliers do to what stats, edit this url to what you want to see.

altitudeladder.com/fantasy.php
?mode=ball_6v6
&kills=1
&assists=0.4
&deaths=-0.6
&carrier=1
&xp=0.01
&goals=1
&goalassists=0.5
&receptions=1
&completepasses=1
&recoveries=0

Any values left out default to the current formula (which for ball is all zeros). Values allow decimals, and may be negative or zero.

Thanks for the data!

As you can see in both of the ball charts, there is actually a low amount of randas. Maybe we could up goal worth to 1.2 and stick an average time of possession stat in there.

[Aki1024]

Originally Posted by Disilord

Maybe we could up goal worth to 1.2 and stick an average time of possession stat in there.

The formula quoted is me just pushing numbers as an example without considering what the values actually could/should be. For instance, I feel recoveries should be considered, but needed a zero for the example.

If you want to see what 1.2 on goals does to the fantasy, edit goals from 1 to 1.2. I have a hard time believing that three kills leading to a goal is worth more then a goal though. That's like sinking a three pointer, but the two passes leading to the shot counted for more in basketball. The top scorer is at 1.8 goals per match. The top killer is at 53.2 kills per 10m. About 2 points for the best goal scorer, compared to about 53 points for the best killer? This point span is too wide.

Offense aimed formula
http://www.altitudeladder.com/fantas...recoveries=0.3

Defense aimed formula
http://www.altitudeladder.com/fantas...recoveries=0.2

I assume receptions are any turnover picked up by the player. Is a recovery any same team pickup of a ball (even a received pass)? Does completed passes only count toward the passer or does the receiver also gain this stat?

I can see from adding these formulas that I need a displayed readable formula on the page to be copy pasted.

[stevemp12]

I think +/- like in hockey would be a nice stat.

[Aki1024]

Originally Posted by stevemp12

I think +/- like in hockey would be a nice stat.

In hockey, it is well defined when the team has fewer players on the ice. In ball, we can't know if while a goal was scored if a player was dead. Assuming your suggesting just watching the end score to determine +/-, then this is do able.

Just to be sure I understand +/- correctly, a game ending in 6 to 3 is +3 for the winning team and -3 for the losing team? Would there be a multiplier on top of this per match stat?

[stevemp12]

right. it would be total team goals scored-total team goals against. 1pt. per +x, and -1 pt for -x.

[Disilord]

Originally Posted by Aki1024

The formula quoted is me just pushing numbers as an example without considering what the values actually could/should be. For instance, I feel recoveries should be considered, but needed a zero for the example.

If you want to see what 1.2 on goals does to the fantasy, edit goals from 1 to 1.2. I have a hard time believing that three kills leading to a goal is worth more then a goal though. That's like sinking a three pointer, but the two passes leading to the shot counted for more in basketball. The top scorer is at 1.8 goals per match. The top killer is at 53.2 kills per 10m. About 2 points for the best goal scorer, compared to about 53 points for the best killer? This point span is too wide.

Offense aimed formula
http://www.altitudeladder.com/fantas...recoveries=0.3

Defense aimed formula
http://www.altitudeladder.com/fantas...recoveries=0.2

I assume receptions are any turnover picked up by the player. Is a recovery any same team pickup of a ball (even a received pass)? Does completed passes only count toward the passer or does the receiver also gain this stat?

I can see from adding these formulas that I need a displayed readable formula on the page to be copy pasted.

Can you post the formulas for what you did with these 2 formulas? Thanks

Thats better for randas, but are we to balance out the offensive formula with the defensive formula or not?

And are we going to have the implication of a midfield formula for loopies? I know loopies are an important part of the game as they can both kill and score which are needed the qualities of a good loopy. For instance threevenge, as you can see he is good in offensive stats because of his offensive abilities are good, but when it comes to defense, they suck. Or we could have 1 slot be for defensive loopy and offensive loopy?

How many slots are we thinking for a team?

Thanks

[Threevenge]

Originally Posted by Disilord

For instance threevenge, as you can see he is good in offensive stats because of his offensive abilities are good, but when it comes to defense, they suck.

QFST (Quoted For Slight Truth) :P

I'm unsure if I'd say my defensive abilities suck, just that I don't kill when I'm on defense: I disrupt with emps, guard passing lanes and steal passes, etc. It's the usual way a lot of talented loopies handle defense. Just doesn't contribute much to the stats other than maybe kill assists.

You don't have to kill to win, though it helps. All that matters is that little spiky ball goes into your opponent's goal 6 times.

[undertheinfluence]

With the crazy ladder teams one can get, the often faulty stats, and the chance of certain players not playing as often as others... It seems impossible for this to be accurate especially if its solely based on either averages or totals from the ball stats... This may have been discussed in the tbd fantasy thread but im not sure.

[Disilord]

Originally Posted by Threevenge

QFST (Quoted For Slight Truth) :P

I'm unsure if I'd say my defensive abilities suck, just that I don't kill when I'm on defense: I disrupt with emps, guard passing lanes and steal passes, etc. It's the usual way a lot of talented loopies handle defense. Just doesn't contribute much to the stats other than maybe kill assists.

You don't have to kill to win, though it helps. All that matters is that little spiky ball goes into your opponent's goal 6 times.

Threevenge I wasn't meaning to be any offebsive to you, sorry if I did, I was just using you as an example between the two different formulas.

Yea, that's what I was trying to portray with loopies, it's like we need a separate stat for them or something along those lines.

[Disilord]

Originally Posted by undertheinfluence

With the crazy ladder teams one can get, the often faulty stats, and the chance of certain players not playing as often as others... It seems impossible for this to be accurate especially if its solely based on either averages or totals from the ball stats... This may have been discussed in the tbd fantasy thread but im not sure.

I'm not sure if you have played fantasy football before but this is what it is all about. Not every player is going to have the same rating every week. Players can play bad one week and be great the next. That's what makes fantasy sports so fun. It's a virtual fantasy.

For example, Maurice Jones Drew was one of the top picks for RBs this year for fantasy football. I picked him as one as my top picks and guess how he did? Terrible. Until the 13th week of the season were he became hot.

Overall, this is for fun. It's just a game.

[Threevenge]

Originally Posted by Disilord

Threevenge I wasn't meaning to be any offebsive to you, sorry if I did, I was just using you as an example between the two different formulas.

No offense taken, just thought it was an interesting point.

[Tekn0]

narf! the offense formula is all wrong because it has me in the top 10

[mjolnir416]

I think that stats like recoveries and ball carrier kills are important. My philosophy is they cant score if you have the ball. Obviously, there is more to it than that, but that is what it boils down to imo.

[mjolnir416]

Originally Posted by Aki1024

I assume receptions are any turnover picked up by the player. Is a recovery any same team pickup of a ball (even a received pass)? Does completed passes only count toward the passer or does the receiver also gain this stat?

I believe its, recptions are when you a ball that your team shoots, but not drops from a death, recovery is when you pick up the other teams ball, possibly also from your team death drops, completed passes only to to the player shooting the ball being gotten by a player from your team. Receiver would get a reception credit not a completed pass credit.

[Dark_Sage]

I'd like to point out that in one formula I'm 2nd and in the other I'm 15th, so obviously I'm awesome. (though I think it needs to be a combination of both, with heavies and lights semi-evenly distributed).

[Disilord]

In the end, I think this boils down to which formula the league decides on which is chosen by the team managers by a vote.

[Flight 666]

Who cares about ball. jez.

[Ribilla]

I've had a crack at this.

I tried to balance the formulas (so they had a similar max and min score) then combine them to get a mix of offence and defence, but it just ended up with loopies owning everything. After fiddling the numbers, I've come up with a formula where all the plane types are pretty equally dispersed and the top 25 looks about right, even if the top 10 doesn't. Also Jewbag ends up right at the bottom, which is nice.

Problems are that it benefits whores too much (see me and aya) because it punishes too heavily for deaths, but this was necessary to get the loopies away from the top and also that there are some good players (blln, elmo, jimlad) who are much lower than they should be. Basically to do well you have to either have a good ratio or score a lot. Players such as donk who die very infrequently, but aren't the best runners in the game benefit too much from the high death penalty. Getting rid of the bug where everyone gets extra deaths would stop a lot of the scores being so negative. Also the top 10 are significantly more valuable than everyone below them, XX2 in particular is best statwhore and I don't like how players like War Machine are above Mattman

http://www.altitudeladder.com/fantas...recoveries=0.6

Kills 0.8
Assists 0.2
Deaths -3
Carrier 4.5
XP 0
Goals 20
Goal Assists 15
Receptions 4
Passes 2
Recoveries 0.6

With some fiddling I think we can have a single formula that works.

[elxir]

top 5 in both modes ^^

[Ribilla]

Originally Posted by elxir

top 5 in both modes ^^

Make sure you order it by score and not rank.

[elxir]

999's don't count they don't have enough playtime to be eligible :O

[Threevenge]

Yeah, I don't like the results of your formula Rib. It's a decent idea but in no situation should there be that many players with negative values.

I believe our best course of action is to get a vote soon on whether we want to take the formula approach or go with a rotisserie approach (or perhaps another solution that appears).

[Ribilla]

Originally Posted by Threevenge

Yeah, I don't like the results of your formula Rib. It's a decent idea but in no situation should there be that many players with negative values.

I believe our best course of action is to get a vote soon on whether we want to take the formula approach or go with a rotisserie approach (or perhaps another solution that appears).

It shouldn't be too hard to increase the score of everyone, but why do you think it matters? Fantasy leagues are only ever relative.

[Aki1024]

A negative score suggest that the player is a detriment to the team.

[Ribilla]

Originally Posted by Aki1024

A negative score suggest that the player is a detriment to the team.

Your formulas also give a negative score. Can you not just whack a little code in so that the column labelled 'Fantasy Score' actually takes the totals from the columns + 100, so no one kills themselves that they've got a negative number.

[elxir]

Originally Posted by Aki1024

A negative score suggest that the player is a detriment to the team.

well...that seems painfully accurate, lol

like, after a quick review of the negative rank players, it's everyone who is either god awful at ball or playing not-their-main-plane and therefore think they are good when really they suck a lot of d...

[Disilord]

I don't think we should have 1 main formula determiner. I think the system should have 2 separate formulas to determine fantasy points. In fantasy football, no 2 different positions have the same formula. This would be hard to do, creating multiple formulas that balance each other out. But I think it is the best way. One way to help implement this would be to create a unique team manager system.
Offensive:
Loopy
Randa
Randa
Etc...

Defensive:
Whale
Whale
Etc...