1911 replies to this topic

[Wintersdark]

carl kerensky, on 10 July 2014 - 08:24 AM, said:

Some stats would help here but it seems to me that most of my deaths are due to engine destruction through the CT, even when running XL engines. Currently engine health I believe is 15 points. If this were to be increased to say 50 points (it stands to reason a fusion reactor would be heavily shielded) then TTK would be greatly increased without changing any other fundamentals about the mech. I think you would take more crits to other systems (HS, weapons etc) before dying.

This would increase TTK (no more one shots) and also increase the immersion factor by having your mech fall apart around you more. As an added immersion and balancing factor, as your engine take more damage you can suffer other set backs like reduced speed, higher heat etc.

Carl, a heads up on game mechanics.

"Engine Destruction" only happens when the side torso IS XL; both side torsos Clan XL; CT any mech are destroyed. That's just how it's labeled.

Otherwise, engines can and to take damage/crits, but when reduced to zero health nothing happens.

You can not kill a mech via crits to the engine. To test: Take a 4+ MG mech to the Training Grounds, and try to CT core an Atlas. You'd blow up the engine loooong before cutting through it's CT if that was possible, but no. You'll ALWAYS kill him via removing his CT entirely.

It's been this way at least since the start of open beta in Nov 2012.

Void Angel, on 15 July 2014 - 10:51 PM, said:

Er, but Engine destruction happens when the component is destroyed by having its internal structure depleted, right - not when the engine criticals take damage?

PS, we wuv you.

Correct.

[Void Angel]

I know, I was excercising indirection as a form of courtesy.

[Karl Berg]

Cimarb, on 10 July 2014 - 04:20 PM, said:

Totally a matter of phrasing. Elo goes up when you WIN despite the prediction, or LOSE despite the same thing. It is still determined by your W/L stat, but dependent on the predicted outcome. Very minor difference.

Hey Cimarb. It's true somewhat that Elo is related to your win/loss ratio. It adjusts based on your predicted chance to win however, and it turns out that this is a really important property. By adjusting in the manner proposed in Elo's algorithm, the system actually converges over time to a value that is not dependent on your win/loss ratio.

Simply put, if your predicted chance to win a game is 100%, and you actually win, your Elo doesn't increase as a result. Put another way, your Elo changes significantly only when you surprise the matchmaker.

The matchmaker is actually training itself. It's making predictions about whether or not you are going to win or loose a game beforehand, and depending on the actual outcome of the game, it's adjusting it's estimation of your skill.

[Karl Berg]

Void Angel, on 15 July 2014 - 10:51 PM, said:

Er, but Engine destruction happens when the component is destroyed by having its internal structure depleted, right - not when the engine criticals take damage?

PS, we wuv you.

Heh, honestly I'm not sure whether or not engine health is a factor in engine destruction. I know some significant changes were made to internals health recently with the release of clan mechs.

By the way, someone asked about mech heights and scaling with the whole foot per second HUD fall indicator. I talked to the gameplay guys about this recently, and it turns out the display units were simply mislabeled due to a miss-communication.

[Void Angel]

Interesting. I'd honestly never considered whether or not the 'Mechs were to scale in that way.

Apropos of nothing, I'm quite happy with the persistent settings of the Play Now button. Those extra clicks are annoying - now I'm just hoping we can get the "public/private" match setting rolled into the drop-down. Bu in any case, that was one of the little touches that really helps the flow of gameplay and adds polish to the experience of playing the game.

[ArkahnX]

About increasing engine health, I just want to chime in to say that I have recently mailed support, and they have said that the internal components do not add to your internal health, and their health is strictly for determining whether it will be destroyed in a crit.

[Cimarb]

Coralld, on 15 July 2014 - 05:38 PM, said:

As Void Angel pointed out, the EI implants are canon, and with the fantastic art team PGI has at their disposal, I am sure they can make them appear way less cartoony.

Also, its your own customizable character, so if you don't want those then fine, don't put them on.

It would also be amazing if you can give your character a flight jacket with your house/merc/Clan affiliation logo on the back of it. (This right here is what I want most.)

My issue is I DO want the EI "tattoos", but would not want them to glow, so if they glow, it stops me from using them the way I would like (and which fits canon much more than glowing anime tats).

I am extremely excited about finally getting some sort of customization like this, though, even of it has nothing to do with gameplay. Just being able to have an identity (beyond a name) is huge for immersion!

Karl Berg, on 15 July 2014 - 11:03 PM, said:

Hey Cimarb. It's true somewhat that Elo is related to your win/loss ratio. It adjusts based on your predicted chance to win however, and it turns out that this is a really important property. By adjusting in the manner proposed in Elo's algorithm, the system actually converges over time to a value that is not dependent on your win/loss ratio.

Simply put, if your predicted chance to win a game is 100%, and you actually win, your Elo doesn't increase as a result. Put another way, your Elo changes significantly only when you surprise the matchmaker.

The matchmaker is actually training itself. It's making predictions about whether or not you are going to win or loose a game beforehand, and depending on the actual outcome of the game, it's adjusting it's estimation of your skill.

That is very true, and it may just be the changes you guys have been fiddling with lately, but to be blunt the matches seem horribly....random.... I will have brand new people, Founders, tournament winners and people that could not fight there way out of a wet paper bag, and everything in between, all in the same match. Whether it is because our Elo is averaging out to be the same or not matters little to me, as the brand new people will have a horrid experience getting stomped, the veteran players will have a horrid experience as their team disintegrates around them, and it will be a lopsided stomp depending on who lands where.

I realize that it will all average out, but we are not talking just about numbers - we are talking about quality of life (or death) here. If I lose 50 stomps, and win 50 stomps, it is perfectly averaged and "working as intended", but in my eyes I have lost 100 matches to the Elo-random Overlord. I want GOOD matches, not "averaged-over-five-months-to-even-out" matches.

Does that make sense?

[Coralld]

Cimarb, on 16 July 2014 - 07:05 AM, said:

My issue is I DO want the EI "tattoos", but would not want them to glow, so if they glow, it stops me from using them the way I would like (and which fits canon much more than glowing anime tats).

I am extremely excited about finally getting some sort of customization like this, though, even of it has nothing to do with gameplay. Just being able to have an identity (beyond a name) is huge for immersion!

Oh, you just have an issue with the "glowing" part of my statement. I just said "glowing tats" like in the animated show so people knew what it was I was talking about. Not everyone knows what EI implants are and how they would look (kinda).
They wouldn't have to glow at all if PGI put in the EI implants in the character creator.

[Kyle Reece]

Karl, not wanting to put words in your mouth but surely in regards to how the MWO implementation of Elo works wouldn't that really be...

Karl Berg, on 15 July 2014 - 11:03 PM, said:

Simply put, if your [teams] predicted chance to win a game is 100%, and you actually win, your [personal] Elo doesn't increase as a result. Put another way, your [personal] Elo changes significantly only when you [your team] surprise the matchmaker.

Or have I interpreted the various CC and dev posts incorrectly?

[Cimarb]

Kyle Reece, on 16 July 2014 - 09:51 AM, said:

Karl, not wanting to put words in your mouth but surely in regards to how the MWO implementation of Elo works wouldn't that really be...

Or have I interpreted the various CC and dev posts incorrectly?

Your edits are exactly why I dislike this system... Having my personal "skill" adjusted based upon team performance irritates me, but I have already argued that way too much for this thread...

[Void Angel]

It's a team game; your personal rating is adjusted in exact proportion to your expected contribution to that team.

[Cimarb]

Void Angel, on 16 July 2014 - 03:51 PM, said:

It's a team game; your personal rating is adjusted in exact proportion to your expected contribution to that team.

Your job is a 'team effort', so should your compensation be dependent on the teams success, no matter how much work you did yourself? While in a perfect world that would work, we all know how poorly it works in practice. Putting in 80-hour weeks and getting the same pay as everyone on welfare, well, sucks.

[Troggy]

@ Karl Berg, et. al.

It seems to me that if an increase in TTK is desired, there is a relatively simple solution that has essentially no risk of unbalancing the game (in fact it would allow a powerful chassis balancing system):

Allow (some) mechs to carry more armour (especially mechs that are currently "bad"). Not lighter, or tougher armour, mind you. More armour. This will have a 2-fold effect on time to kill. 1) Mechs will be tougher, 2) Mechs will have less guns, because they have more weight of armour.

Currently, nobody ever restricts their armour on any engine-containing component, and no light or medium restricts leg armour. Right now the guns/armour balance is weighted heavily towards guns. I.e. virtually, all mechs can carry too many guns for heat capacity or good sense. None carry "too much" armour.

If the Awesome, for example, had 50% more CT max armour (and corresponding Int. HP) it might actually see some use. A fast tank-y assault might have a role on the battlefield (which is actually canonical for the Awesome), even with a subpar weapons mix. Likewise, for the Dragon - it would fulfill its purpose of a fast skirmisher must more effectively with 30% more armour on each of its torso components (even at the cost of a little weaponry). Likewise, Hunchback - the added armour would compensate for its low speed, lack of JJs, and relatively short-range weapons mix.

One of the largest problems currently is that heavier mechs within a class are universally better because they are more survivable AND more heavily armed (shorter TTK). Focusing on boosting the armour on some of the lighter, faster options within the mech classes, would encourage their use, and increase time-to-kill without any substantial re-work of the game mechanics.

Finally, if people don't like the change, they can always carry less armour.
--
Troggy

Edited by Troggy, 16 July 2014 - 04:39 PM.

[Void Angel]

Cimarb, on 16 July 2014 - 04:22 PM, said:

Your job is a 'team effort', so should your compensation be dependent on the teams success, no matter how much work you did yourself? While in a perfect world that would work, we all know how poorly it works in practice. Putting in 80-hour weeks and getting the same pay as everyone on welfare, well, sucks.

Except that's not what's happening. What's happening is that you're being randomly matched in a pickup game with other players against another randomly matched team, and your personal ranking is adjusted based on how your team, including you, is expected to perform against the other team. This is not unfair in the least - what would be unfair would be to punish you by dropping your Elo the full amount when your lower-ranked teammates drop the ball against a higher-rated enemy team.

Elo works, it's not unfair, there is no better way to do it, and no number of bad analogies will change these facts.

Edited by Void Angel, 16 July 2014 - 04:49 PM.

[Cimarb]

Void Angel, on 16 July 2014 - 04:48 PM, said:

Except that's not what's happening. What's happening is that you're being randomly matched in a pickup game with other players against another randomly matched team, and your personal ranking is adjusted based on how your team, including you, is expected to perform against the other team. This is not unfair in the least - what would be unfair would be to punish you by dropping your Elo the full amount when your lower-ranked teammates drop the ball against a higher-rated enemy team.

Elo works, it's not unfair, there is no better way to do it, and no number of bad analogies will change these facts.

*in your opinion.

I happen to not agree, but I am in the minority, so it is what it is.

[Void Angel]

Hard math has never been "opinion."

[Heffay]

Cimarb, on 16 July 2014 - 05:06 PM, said:

*in your opinion.

I happen to not agree, but I am in the minority, so it is what it is.

That's not an opinion. That is exactly what is happening.

[Hex Pallett]

Anywhere can I actually take a look at the ELO algorithm?

[Wintersdark]

Helmstif, on 16 July 2014 - 09:37 PM, said:

Anywhere can I actually take a look at the ELO algorithm?

A good starting point for Elo info is here: http://en.wikipedia...._rating_system. There are links in the references regarding Elo in team sports as well.

As to our specific implementation, the formula was posted... Crap. It was posted somewhere on the forums here, but I'm not sure exactly where. I suspect within this thread, fairly early on (Elo is a common discussion point here), but it may well have been elsewhere. Did you check Paul's command chair posts in regards to it?

[Kageru Ikazuchi]

Helmstif, on 16 July 2014 - 09:37 PM, said:

Anywhere can I actually take a look at the ELO algorithm?

> Here < is the Command Chair post you're looking for.