New Scores for DD/JD MVP are live!

[Steadman]

The old way of scoring jav and wb that had MVP decided by kills with a tiebreaker on deaths is now long gone. The new formula Biet created is an improvement on his old formula, which now takes into accounts the usual k/d but now also includes knockouts, staying alive, and teamkills (where applicable).

For those interested, the new formula is:

(Kills*2) + Knockouts + Staying Alive - Deaths -Teamkills

What does this mean for the the future of TWD?
Getting MVP will mean putting an emphasis on winning rather than playing for kills at the detriment of your own team. Maybe it isn't worth it to shoot that 2fer killing your teammate. While killing your opponents is still an important aspect of being the game's MVP, making team plays, avoiding detrimental shots, and winning will now be major aspects of being considered MVP of the game.

Keep in mind that some of the values listed above may be tweaked as time goes on.

[zidane]

adorable font man, surprised there aren't some flower petals and your signature at the end

[Exalt]

The old way of scoring jav and wb that had MVP decided by kills with a tiebreaker on deaths is now long gone. The new formula Biet created is an improvement on his old formula, which now takes into accounts the usual k/d but now also includes knockouts, staying alive, and teamkills (where applicable).

For those interested, the new formula is:

(Kills*2) + Knockouts + Staying Alive - Deaths -Teamkills

What does this mean for the the future of TWD?
Getting MVP will mean putting an emphasis on winning rather than playing for kills at the detriment of your own team. Maybe it isn't worth it to shoot that 2fer killing your teammate. While killing your opponents is still an important aspect of being the game's MVP, making team plays, avoiding detrimental shots, and winning will now be major aspects of being considered MVP of the game.

Keep in mind that some of the values listed above may be tweaked as time goes on.

This is a good start. If only players themselves weren't too retarded to understand this naturally.

[apt]

thank you but why is this written in comic sans

[beam]

thank you but why is this written in comic sans

cuz the guy writing it is a joke lol

[RageRitual]

Amazing how 4 posts here, none have anything to really do with the topic itself, even though I did chuckle at zidane's post. So many people are quick to trash the rating system, how about actual thoughts on it when it does get work put into it?

[saiyan]

Its an improvement, but a better one would be getting rid of mvp altogether.

Add separate (global) matchmaking ratings to all wb/jav/base arenas and minimize aliasing. You'll incentivize winning instead of playing for stats and finally get a relatively accurate, objective measurement of impactfulness. Watch the level go up across the board in each league as people get obsessed with their MMR (which is good) and start optimizing their game to win. It would look totally different lol... Jessup might stop sitting back and letting his teammates get showered in bullets every match... Or (much more likely) not change a thing and cry about his abysmal MMR nonstop.

Not sure why this wasnt done 10+ years ago tbh.

Edit: Also, just to randomly vent at wbs like jessup.

YOU NEED TO MAKE PEOPLE SHOOT AT YOU (learn to dodge) it makes the game easier for your teammates. If no one is ever shooting at you, you are ruining the game.

Ty

[Steadman]

saiyan Creating an MMR system with the limited amount of stats that we're able to pull would have problems in and of itself. For one thing it might favor players on stacked squads too heavily. It would be a different situation if we had a larger playerbase (1000+ active players) and could track things like damage dealt and assists, but unfortunately those are not options.

I believe Ogron tried an MMR system in TSL, which was a randomized team league from a few years back, and it had issues for the reasons listed above and hasn't been hosted since.

[saiyan]

Steadman Youre right to some extent, its best for wbduel/javduel/?go base, but I believe it could work anywhere.. That TSL implementation was probably different than what I'm suggesting. You wouldnt need to track any stats like assists/dmg. You simply need to win. The bot would average the ratings of each team and develop a chance to win for both. Afterwards you would get a rating change inversely proportional to your chance to win. It would actually disincentivize stacking and incentivize carrying low rated teams in casual matches. Winning on a superstacked team would net u 0 rating while losing would cause a large decrease. On the other hand carring a low rated team would be very lucrative because you'd gain huge amounts for winning and lose close to nothing if you lost.

Start everyone at a baseline, and over a long enough period without alias abuse people will adjust to their true rating. When your rating gets really high, you would have no choice but to carry low rated players if you want to keep progressing.

[Exalt]

Steadman Youre right to some extent, its best for wbduel/javduel/?go base, but I believe it could work anywhere.. That TSL implementation was probably different than what I'm suggesting. You wouldnt need to track any stats like assists/dmg. You simply need to win. The bot would average the ratings of each team and develop a chance to win for both. Afterwards you would get a rating change inversely proportional to your chance to win. It would actually disincentivize stacking and incentivize carrying low rated teams in casual matches. Winning on a superstacked team would net u 0 rating while losing would cause a large decrease. On the other hand carring a low rated team would be very lucrative because you'd gain huge amounts for winning and lose close to nothing if you lost.

Start everyone at a baseline, and over a long enough period without alias abuse people will adjust to their true rating. When your rating gets really high, you would have no choice but to carry low rated players if you want to keep progressing.

This is a good idea honestly, but I don't think it would work out due to the players themselves. This game has almost never seen people NOT stack squads, and as much as I hate to say it, there's a good reason to do so. It's almost impossible to win with certain players if they don't even know the basics of teaming against a squad of higher-skilled players that do, and if only one squad stacks, then they're almost certainly the favorites to win something like TWL. This might work for TWDT though, where people are forced to spread the talent around.

[midoent]

Steadman Youre right to some extent, its best for wbduel/javduel/?go base, but I believe it could work anywhere.. That TSL implementation was probably different than what I'm suggesting. You wouldnt need to track any stats like assists/dmg. You simply need to win. The bot would average the ratings of each team and develop a chance to win for both. Afterwards you would get a rating change inversely proportional to your chance to win. It would actually disincentivize stacking and incentivize carrying low rated teams in casual matches. Winning on a superstacked team would net u 0 rating while losing would cause a large decrease. On the other hand carring a low rated team would be very lucrative because you'd gain huge amounts for winning and lose close to nothing if you lost.

Start everyone at a baseline, and over a long enough period without alias abuse people will adjust to their true rating. When your rating gets really high, you would have no choice but to carry low rated players if you want to keep progressing.

Good idea. The balance for this would be hard to figure out though.