MVP in a timed race game

[Force of Nature]

The problem is I want to define the MVP such that a ship that excells in what it is supposed to do get MVP.

1st condition: Any ship in base should be able to get an mvp.

2nd condition: The ships need to be comparable to each other

So now i'll let all you sleuths think of possible ways to determine a solution given that you hold the two conditions.

I think generally you have to think of how good a ship performs its goals and then map it onto a rating. Given what is important to its function in a base game.

Do this to all the ships and the player with the best rating get MVP. Now ideas on how each ship gets mapped onto the rating please

[Adman]

i think the players should vote on a MVP for base.

Compare spiders, say one has 170 kills and the other has 140 but a higher ave who made more of an impact on the game? Most people agree that the guy with more kills.

Also, it's hard for sharks to get recognized for MVP because they are a tandem, but games can be won by sharks.

For terr, it's easier, you can see clearly if the terrier won the game for the team, extremely losw deaths, or high bounties amke it clear, but for TWL (because lots of people watch the games) the spectators should vote for MVP (not the players) but speccers for the squad's that play could.

Basing is not as cut and dry as the dueking leagues because meh it's just not. the guy who got 16-4 is more important than the guy who went 8-6.

[bloodzombie]

voting is the best way.

skip it completely in ?go base or anything like that, but if it went to twl, voting would make it cool.

[HellzNo!]

ship switches kind of fuck around with mvp.

[Randedl]

points
deaths/kills per minute
terr kills
entrances


prolly have voting though yeah

[Force of Nature]

ok
voting is nice. And its probably a good way. But I want some stat rating wise. When voting people always tend to favor well known people compared to anyone new who has done equally well. And for this aspect I like the rating system better. Granted it is not perfect by any sense but its give something for people to aim for.

So I'm going to phrase the question differently.

For each of the ships what stats do you think are the most valuable. For eg: terr have a good ratio. every death effectively halfs the ratio type thing. This is just an example.

Also tell me For each ship what do you see for when you determine they deserve to be mvp. How do you decide someone deserves to be MVP for each ship that is allowed in basing. For eg: wb lots of terr kills is important.
If you can relate it to stats that would be great however if you cant that would be fine as well. Lots of people have strong beliefs in how one ship in basing should be played so lets hear them or what they think deserves mvp for a person playing that ship.

[Jason]

Here's another vote for voting, and if no one bothers to vote, just let the MVP default to the standard "dude with the most points."

[Jason]

Hmmm... let's see here.

Warbird

Terr kills are probably the most important factor in determing whether or not a basebird did his job well. I have absolutely no idea if it's possible for the bot to detect and record how many terr kills a basebird got though. If you want to simply go by stats, I think a basebird should have at least a 3:1 ratio to be considered for MVP.

Javelin

Terr kills and cram breaking are, in my opinion, the two most important factors in determining whether or not a basejav did his job well. This goes back to the bot being able to detect certain things and record them, but if you could somehow figure out how many terr kills a basejav got, combined with how many 3fers (or higher) said basejav got, I think you'd have a good idea of his rating. I say you should count 3fers or higher because that's what a cram break generally involves. If want to simply go by stats, a basejav should have at least a 3:1 to be considered for MVP.

Spider

This one's easy. Spiders are killing machines. You go purely by stats and look at who has the most kills, or possibly the best ratio. Some will argue that a good spider doesn't necessarily have the most kills because good spiders will meatshield for their terr and what not, but I just don't see how you're ever going to get a bot to realize that a spider was meatshielding.

Terr

This one's pretty easy, too. Look at the ratio and number of deaths. A terr that deserves MVP will have a really low number of deaths, and an even better terr will have at least a 5:1 ratio.

Shark

I have no fucking clue. It's just damn near impossible for the bot to realize that a pair of sharks did a good job trading off in the cram. You can't go by stats because sharks just about always end up with a shitty rec. The only way I see a shark getting MVP is via a vote I'm afraid.


In addition to the specifics I've mentioned, you could always take into consideration things like points, ave, rating and all the built-in stats just for good measure. And I didn't bother mentioning the lanc because I don't think it'll be used more than a few times by some crazy dudes like those Pallies.

[Force of Nature]

You will be amazed what you can do with a bot. Obviously this is a big project if going to be done right. However, it is good documentation for someone who might be willing to do it over a couple of months.

Now, I'm explain the examples why I said you will be amazed:
Terr kills are no problem. That stat was already recorded in the TWL bots when we had TWL games.

Secondly with meat shielding, you can approximate. This involves more work obviously. I think this feature alone would take about 7 days which is a conservative estimate. The way you'd do it is such. You have a radius around the terr which you can get since you have coordinate on the map for the poistion. Remeber the bot gets all the packet data which the server is sending. If the spider ate a bullet (this one is harder since you have to have watch damage on - bots are sysops so you can do it) in that specified radius then you assume it shielded the terr. This is an approximate but still its close.

3fers counting is harder as well since they all show as individual deaths. But you can take the timespan to determine the 3fer so it can be done.

Sharks are obviously harder. Since performing in the cram isn't the only objective. You can treat the sharks as a pair and evaluate and give mvp to both sharks is one option. If this is the case you can time all the instances where the team was sharkless to use in evaluating. You can also count the number of repels before they die which can be another statistic.

On a sidenote. If the modified lanc actually gets implemented into TWD and TWL. In my opinion it is quite a formidable ship. I would also like to see the mini weasel

[za gophar]

i do not think voting is a good solution. it would be too opinionated, and just as inaccurate. people would rarely vote for unknown players who may have gotten the mvp otherwise. the current mvp system is a good way for new players to gain recognition and old ones to continue being rewarded for the efforts to keep their spider game up. anyone good at shark or terr, in my opinion, has to be at least a par spider. they will get their fair share of recognition still. this is another reason why spiders deserve mvp honors; they are the least recognized during the game.

however if the system must be changed i guess i think it would be best for the bot to state multiple honors at the end of each game such as most terr kills, best ratio, highest average kill, highest kills, lowest deaths, yatta yatta yatta.

[Fit of Rage]

voting is the best way.

skip it completely in ?go base or anything like that, but if it went to twl, voting would make it cool.

I agree. However, I think that only players of the participating squads should be able to vote, not people in spec mode, if that's possible to implement.

[Cig Smoke]

however if the system must be changed i guess i think it would be best for the bot to state multiple honors at the end of each game such as most terr kills, best ratio, highest average kill, highest kills, lowest deaths, yatta yatta yatta.

Ya thats how DSB is, it shows all that stuff u said + more. I dont see it as gonna be totatly equal to all ships in basing unless its a vote. That would probably be the best way to get total equalness for each ship. A shark/pair of sharks wont ever get MVP unless spec votes for them. A spider with 180+ kills will always get MVP over 2 sharks, or a terr who is 50-1 will get it to, even if the sharks saved that terr 10+ times.

[Force of Nature]

Instead of telling me what everyone else said, which is to vote. Can we have something constructive like jason did or comments on why his suggestions should be changed. Its nice that you think that people need to vote. But what do people look for when they vote? They must look for something. They don't just randoming point fingers and say that guy deserves MVP. Well some do but thats not the point. Anyway, more about how you think each ship should be judged.

[cundor]

vote it is

[Rinse]

Some zones have a bot which tracks damage delt and damage received. I could prolly whip one up and we could test it. I mean I spider can get 100 kills with alot less damage delt then a spider with 50 kills. If a spider just shoots randomly and gets the last hit then they will obviously get more kills. I think accuracy and damage should play a role in determining the MVP. Also with the terr, they should try to get as little damage as possible and vice versa for sharks.