Go Back   Star Trek Online > Feedback > PvP Gameplay
Login

Reply
 
Thread Tools Display Modes
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 21
03-24-2010, 03:00 PM
in this game it would only work for FvF. there just aren't enough people in the FvK queues to make it workable -- you'd end up with 5 klingons playing the same 15 or so similarly-rated feds all night, while all the rest of the feds sat there with no similarly-rated klingons to fight.

-ken
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 22
03-24-2010, 03:51 PM
Quote:
Originally Posted by wildfyre010 View Post
I'm really not crazy about the idea of being placed in a T4 match with my T3 Cruiser just because I'm winning more T3 matches than most other people. One of the things that's generally true about PvP is that good players do well, and reap the rewards of their success. I have no problem being matched against opponents of similar skill - that is a necessity of any matchmaking system - but I have a big problem with artificial impediments to my success that are introduced because I'm doing too well.

Said another way, my victory or defeat should be a function of my skill and the skill of my enemies and teammates. Adding additional conditions, like escalating my T3 ship into a T4 map because the numbers suggest that's where I'll be balanced, is just another way of placing a handicap on me to even the playing field for inferior players.
First, your rating would be determined by your participation as compared to the participation of everyone else in your match. This is calculated independent of whether you win or lose your match.

I'm sure we have all been in situations where we see (or perform) participation that outmatches everyone else in the match, yet we still lose that match due to poor teamwork. My proposed system takes this situation into account, and rewards participation instead of win:loss. This is, therefore, a function of your skill in relation to the skill of your team and your enemies.

This rating will contribute toward your future victories such that it will attempt to normalise your win:loss ratio to 50% by matching you against those who participate relatively the same as you. In some cases, this might mean higher tier players, but, again, only if they are already participating the same as you. Which means, there is no difficulty imbalance between you and them. There is no impediment to you.

Unless you see normalisation of win:loss to 50% as an impediment. In which case, well, not everyone likes a challenge.
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 23
03-24-2010, 04:01 PM
Quote:
Originally Posted by Snowreap
in this game it would only work for FvF. there just aren't enough people in the FvK queues to make it workable -- you'd end up with 5 klingons playing the same 15 or so similarly-rated feds all night, while all the rest of the feds sat there with no similarly-rated klingons to fight.

-ken
This statement makes several assumptions, which may or may not be true 100% (or even 51%, a majority) of the time:

1) There are only, at best, 5:15 Klingon:Federation (1:3) PVPers at any given time. Or, Klingon PVPers at best represent 25% of the total PVP population at any given time.

2) Klingon PVPers will be of the same rating 100% (or even 51%, a majority) of the time.

I question the veracity of these assumptions. I furthermore purport that my proposed system offers the most challenge to all PVPers while also being equally fair, yet it also takes into account faction population imbalances because of its 'blindness' to tiers. (No pun intended.)

Furthermore, there is freedom of movement between ratings brackets (albeit slightly impeded for backwards or downwards movement), so there is nothing stopping similarly rated players from seeing different opponents over a short period of time (e.g., a night, in your example).
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 24
03-24-2010, 04:37 PM
Quote:
Originally Posted by Matt_Dravis
This statement makes several assumptions, which may or may not be true 100% (or even 51%, a majority) of the time:

1) There are only, at best, 5:15 Klingon:Federation (1:3) PVPers at any given time. Or, Klingon PVPers at best represent 25% of the total PVP population at any given time.

2) Klingon PVPers will be of the same rating 100% (or even 51%, a majority) of the time.

I question the veracity of these assumptions. I furthermore purport that my proposed system offers the most challenge to all PVPers while also being equally fair, yet it also takes into account faction population imbalances because of its 'blindness' to tiers. (No pun intended.)

Furthermore, there is freedom of movement between ratings brackets (albeit slightly impeded for backwards or downwards movement), so there is nothing stopping similarly rated players from seeing different opponents over a short period of time (e.g., a night, in your example).
I think the queue line speaks for itself on this issue dude.
Right now at tier 5 there are over 100 federation in line with less than 5 klingon.
Ok assuming there are 50 per side engaged in various games, that gives us 150-50. That is 3-1 and I am thinking it is more along the lines of 4 or even 5-1.
This is an assumption yes and without knowing how many people are engaged in various battles we have no way of knowing if it is true. However it is still a very safe assumption. Also factor in how long you are in that line.
So yes I think it is safe to assume there are roughly 4-5 times as many federation pvp players than klingon pvp players.
As for overall who knows? I can only go by what I am seeing through the queue line.
At the bottom tier lt6? There were way more klingons and probably still are.
But I don't want to go off about all the tiers, my main concern is what's going on at tier 5.
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 25
03-24-2010, 04:37 PM
they weren't assumptions, they were illustrative examples. my point is, there aren't enough klingons available in FvK queues to do this. the scheme depends on having a large enough population that you can break the players down into rating brackets and match them up, and still have enough players in each bracket to actually start a game.

if we had 5 times as many klingon players as we do now, then this might work for FvK as well. but as low as the klingon population is right now, too often you'll find that the people with the ratings you need simply aren't logged on at that moment. the problem is that low populations magnify the effect of outliers. with higher populations the distribution of players smoothes itself out.

I'm not saying it's a bad idea, I happen to think it's a very good idea. but you need to have enough players in the pool that you're able to find good matches. for the matches to be fair, the distribution of player ratings needs to be the same on both teams. but if there are too few klingons queued, you won't get a good distribution on their side. but you can guarantee that the distribution of players on both sides is the same if you simply use the same population for both sides, i.e. FvF.

here's an extreme example: 5 fed players log on. they are all Lt6, and the worst pvp'ers to have ever lived. there are also 500 other fed players on, of average mean ability with the expected standard normal distribution. on the klingon side, you have 10 players total logged on (it's a slow night). because the klingon population is so low, you don't have a standard normal distribution, instead, 8 of the 10 are above average (because they do nothing but pvp every night, and by now are very good at it). who are you going to match those 5 awful fed players up against so that they will win 50% of the time?

it's a contrived example, of course, but it illustrates the problem when the distribution of ratings isn't the same on both sides. applying this rating scheme to FvF fixes the problem completely, these 5 horrible feds are paired with the 5 next-worst fed players in the fed queues, mixed up into appropximately-equally-rated teams, and they can play each other.

-ken
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 26
03-25-2010, 12:29 AM
Figuring out how to best calculate someone's contributions would be very complicated. The best example I have is in SalvOp, and this is a pretty common scenario for both sides.

Fed team gets a big lead due to the Ks starting out shorthanded, so the Feds get overconfident and decide to stop caring about the flags and just slaughter the outnumbered Ks around the sensor array, leaving one noble and heroic Fed to defend the other 4 flags. Spoiler: Said noble hero was me. If I was a Klingon, boisterous songs would be sung of my deeds that day... but enough about me (for now).

So while every other Fed is racking up damage and points, and even buffing and healing each other, I circle around the starbase for several uneventful minutes keeping a close eye on the 4 undefended blue flags. Of course, a couple smart Ks decide to stop getting slaughtered at sensor array and decide to make a play for the undefended flags. As everyone knows, no lead is safe in SalvOp, the momentum can totally change in under a minute. So I jaunted back and forth between the 4 flags, making sure that white didn't turn into red, barely getting in a couple phaser blasts before the Ks cloaked. We won that contest, but if you based everything off of stats (all stats included, not not just damage), you'd look at the scoreboard and assume I got in one skirmish and then spent the rest of the game parked in front of the Fed spawn point.

Now I personally don't play to top the scoreboard, I'm cool with being unsung support (the songs are still being written), but if we brought in a definitive number to judge PVPers it would leave a lot of stuff out. Just food for thought.
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 27
03-25-2010, 01:39 AM
Quote:
Originally Posted by forsakenlight View Post
This is an assumption yes and without knowing how many people are engaged in various battles we have no way of knowing if it is true. However it is still a very safe assumption. Also factor in how long you are in that line.
So yes I think it is safe to assume there are roughly 4-5 times as many federation pvp players than klingon pvp players.
As for overall who knows? I can only go by what I am seeing through the queue line.
At the bottom tier lt6? There were way more klingons and probably still are.
But I don't want to go off about all the tiers, my main concern is what's going on at tier 5.
Well, this assumes that the PVPers at each tier are constrained to their tiers. My proposed system replaces tier constraints with ratings brackets, so tier-specific population imbalances will not be as great a concern.

Regarding overall population imbalances, see below.
Quote:
Originally Posted by Snowreap
the scheme depends on having a large enough population that you can break the players down into rating brackets and match them up, and still have enough players in each bracket to actually start a game.
These two posts address a fundamental issue with any matchmaking system, which involves population.

According to my proposed system, normalisation to 50% win:loss does require a large population. However, matchmaking in general -- and the operation of the ratings system -- does not require any particular population (above 10 total, of course, if 5v5 is our most basic unit).

More accurately, we do not need to constrain players to their ratings brackets.

Matchmaking will first select from within a ratings bracket, and then move outward if there are not enough players available within that single bracket.

While it is true that there may be situations where PVPers from one bracket might be matched up with and against PVPers of other brackets, this will happen only if it becomes necessary to create PVP matches (i.e., quality-of-service oriented, instead of best-effort oriented). In this scenario, ratings are still calculated normally and largely unaffected, as long as we also apply weights to the different brackets.

For example, a scenario where a PVPer from Bracket 3 is forced to play with and against PVPers from Bracket 1. Without bracket weightings, the participation calculated would be extremely high for the player from Bracket 3, which would temporarily inflate that player's rating. To compensate, we can then apply weightings in the form of a (fractional) coefficient applied to the participation component, where the coefficient represents a ratio of the brackets' average participation.

===

So the participation formula becomes:
Participation = (AvgParticipationOfHighBracket / AvgParticipationOfMyBracket) * (myParticipation / totalParticipation)

And the ratings (accumulation) formula becomes:
myNewRating += (AverageRatingsOfPlayersInGame / 100) * (Participation)

Note that PVP match rewards, and the score screen, will still be based upon (myParticipation / totalParticipation).

===

By applying the bracket coefficient to the participation component and not the entire ratings formula, players from lower brackets are rewarded slightly for competing against players from higher brackets, while the ratings calculation (in that situation of mixed brackets) would be largely unaffected for the players from higher brackets.

===

How does this sound?

I have edited the original post to reflect this feedback. Take a look and let me know. And thanks again for the excellent feedback.
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 28
03-25-2010, 01:43 AM
Quote:
Originally Posted by PSIRockin
Figuring out how to best calculate someone's contributions would be very complicated. The best example I have is in SalvOp, and this is a pretty common scenario for both sides.
As you can see from the original post, we take into account virtually every action from every PVPer in a single match.

This applies best to arena (deathmatch). I am still unsure how to include objective-based participation, such as in capture-and-hold maps. My first reaction is to ignore objective-based scores for participation and ratings calculations, but keep objective-based scores as part of the reward calculation at end of match.
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 29
03-25-2010, 08:51 AM
OP has my full support.
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 30
03-25-2010, 01:55 PM
Quote:
Originally Posted by Matt_Dravis
I am still unsure how to include objective-based participation, such as in capture-and-hold maps. My first reaction is to ignore objective-based scores for participation and ratings calculations, but keep objective-based scores as part of the reward calculation at end of match.
If you're a member of a "good" capture and hold team, your role may be to ninja unguarded points or to stand guard at one that's not seeing heavy fighting in order to prevent it from being stolen. Strategically smart play should be rewarded, not punished.

One mechanism for doing this could be to have all scoring areas on a map "pulse" contribution points to people who are near them. Contribution per pulse could be something like:

Tier modifier * 1000 * number of points held by your team / number of players at that point

This would make it more lucrative to guard flags when your team is ahead (i.e. strategy favors a defensive posture), but not to pile defenders at "cold" points (because of the splitting).

Flags also pulse points during combat, so a solo defender fighting off a ninja would continue receiving flag points in addition to the (relatively sparse) combat contribution. If an area gets "hot", the flag bonus becomes less and less significant since more people are present - but you'd be getting combat contributions instead.

Flags would pulse points to both friendly and attacking players, so sitting at an unguarded point while it caps gives you contribution points as well. Attacking a lightly-defended point would similarly give you contribution points (you'd be splitting them 50/50 with a solo defender until one of you died, of course).

Finally, the system rewards player who fight near flags in general. It pulls action toward the important areas of the map and slightly penalizes people who chase way outside the normal map area - that kind of combat generally contributes less to your side's victory, after all. Arguably, if your team was winning, then you might be doing something useful by pulling enemies away from the points, but you'd be getting combat points for your effort most likely, and it's pretty much indistinguishable whether you're in the middle of nowhere because you're really smart or really dumb.
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off


All times are GMT -7. The time now is 05:55 PM.