What was going on here?

VizMantis

So I just joined the Deadpool pvp, I've got about 100 pts. I get a pop up saying I'd been defeated. The attacker had 1280 points in the pvp and his attacking team consisted of 2 level 550 5*s. The weird thing is, the defeat was worth +0 points to the attacker (-0 for me). What was the point? Just the victory reward?

And why was I getting queued up against such a strong player anyhow? I have no 5*s above 270 and my highest 4* is 285.

amusingfoo1

He was probably farming for ISO. And he got one point, fwiw.

fmftint

VizMantis wrote:

So I just joined the Deadpool pvp, I've got about 100 pts. I get a pop up saying I'd been defeated. The attacker had 1280 points in the pvp and his attacking team consisted of 2 level 550 5*s. The weird thing is, the defeat was worth +0 points to the attacker (-0 for me). What was the point? Just the victory reward?

And why was I getting queued up against such a strong player anyhow? I have no 5*s above 270 and my highest 4* is 285.

Grill reset

HaywireII

I'll do you one better. Two events back I got hit by someones Charlie Angels for a 0 point loss. You bet that I took the 72 point retaliation. I'm betting that he was shielded by that time though.

VizMantis

fmftint wrote:

Grill reset

So you're saying he had a weak team of defenders out for other players to take advantage of and went against my low-risk team to get his strong defenders back out? That'd make more sense than iso farming I think. Although I don't totally understand what all grilling is. I understand it's the baking replacement, but somehow only benefits the super-high level players?

VizMantis

HaywireII wrote:

I'll do you one better. Two events back I got hit by someones Charlie Angels for a 0 point loss. You bet that I took the 72 point retaliation. I'm betting that he was shielded by that time though.

I was thinking I might do that, but I don't have a lot of confidence in my ability to take out 550 black bolt/black widow + 412 deadpool.

Phumade

Its a weird side affect of how the match making works. I'd have to know the exact time of the matches, but you can guess that the following situation occurred.

1. He's probably got one of the higher scores in the entire bracket and he's probably well over 2 std deviation over the avg score.
2. Because no other equivalent rosters with a similar score are actively climbing or unshielded, the game expands the possible opponent que. Past certain thresholds, the game will add 4*, 3*, 2*, 1* class rosters to the opponent mix.

Based on the reported score differentials, he was probably seeing you as a 1 or 2 pt match. You wouldn't believe how common this actually is for 5* players who start events early.

While its hard to guess actual motives,
--snipers generally don't hit below their scored and it would be absurd for a 1200 sniper to hit a 100 pt floater.
--While its still feasible to farm low level teams for iso and covers. Shield intercepts have mostly eliminated the need to farm iso. Now when people farm, its because they are looking for a casual mode to kill time. i.e. endless tetris. For the most part, people don't farm pvp events for iso. Farming is best when your playing a limited timeframe like LRs. Otherwise, your really just playing to kill time.
--More likely, the player either need to grill or reset off a lower level seed class team.

On the bright side, you got a AAA level retal node that will not diminish in value. You can be sure it will be worth 75pts now and 75 pts at the end of the event. I would be licking my lips if another 5* player was dumb enough to give me a retal node from 1200 pts while I was floating at 100

VizMantis

Always interesting to hear about life at the top. Thanks for the detailed analysis Phumade!

JamieMadrox

Late event at really high points you see all kinds of weird teams. Sometimes you'll even see a seed team pop up. It's weird, but a side effect of the system being pushed beyond what it was coded to do by players using good communication.

FaustianDeal

That said, I would disagree with the assertion that farming is only done "to kill time". The other reason why players will hit those teams is because they are trying to grow points in the shard.

Let's assume "farming ISO"
5* characters are incredibly demanding of ISO. Championing a 5* character has the same ISO cost as championing 5x 3* characters. It might not seem like a lot of ISO, but every victory counts, and the power differential between max champed 5s and 3*s or 4*s is so big that there was little actual risk - so it is productive in that regard.

Now let's assume "farming points"
As players get close to the top end of the scores in a given end time (not a bracket, match-making looks across the entire end-time) the algorithm that picks opponents runs out of what it deems to be 'appropriate' options and it starts loading anything it can find. This is sometimes referred to as "breaking MMR". This is not a fixed score, it starts low in events if you start early, and climbs ever higher. Early in events you can find yourself presented with those Qs as early as 300 points. Later in events you could need to be well over 2000 points to be offered those targets.

Some of this requires an understanding of how points are gained and lost as you play. There is going to be a lot here, hopefully some of it is useful or entertaining (otherwise I got out the tinfoil hat and got a furlough from the loony bin for nothing).

There is a formula to determine what an opponent is worth to you, but an oversimplified way to explain it is that hitting someone with your exact score is worth 38 points. People with more points than you are worth more points... they are, roughly speaking, worth +1 point as a target for every +20 points they have in score differential over you. So if they have +100 points over you they should be worth +43.. up to a maximum of 75. That formula for how many points an attacker wins in PVP works all the way up and down the scoreboard.

When you look at how many points the defender loses though, there is another modifier in place. Loses are pro-rated based on how far you are below 1000. It used to be a simple scaled reduction. If you have 500 points, and someone who has ~460 points attacks you - they should receive +40 points. You should lose +40, but you were only @500 you actually only get penalized to the tune of -20. People trying to create points early enough in an event can take advantage of this loss modifier to make points. I hit you for +40, you lose -10, trade a few times - our scores both inflate a little each time. We get to the point where the attacker gets +40 and the target loses -20, then -30... eventually as you get too close to 1000 it becomes zero sum. From that point forward there are no 'new' points (unless someone buys a shield, but then there is no more trading).

As a digression, for education, if the attacking player loses there is an additional 1/3 factor added... so if if they were to lose the hypothetical +40 match we are talking about here... they lose -40... but their target doesn't get +40... the target only gets +13 (40 * 1/3). This modifier was added when players were exploiting the proration factor in order to quickly inflate scores. Trading retaliations to build points takes a while - but back in the day - if you could get 10-20 people to collude and retreat to one another you could get everyone up to about 800 just by sacrificing the bottom half of your roster and using the loaner feature. You weren't using those characters anyway. One additional note of flavor should be mentioned here - players who intentionally tank their scores to 'snipe' or to 'enforce' are now creating a downward pressure on scoring in an entire slice. Points are being destroyed in that process (the act of dumping) and the economy of the slice suffers when it is systematically abused.

But back to topic... the already high-scoring player who is staying out and farming +1 point matches to build their score from 1200 -> 1400 has done more than you might initially think (more than getting another +200 for themselves). The target that they hit probably lost -1 or even 0 points... so they felt no large loss. Even if the attacking player got +6, the odds are good that the target still only lost -2. So their impact to those lower players during the farming process is minimal. But the effect on the whole slice is larger. Let's say I am @1000, and I have this farming player Qd to hit. If that player had shielded @1200 and I hit them from 1000, I will get +48 points. But if they farm to 1400 before shielding, when I hit them I will now get +58. That might not seem substantial initially, but in a coordinated effort I won't be the only person benefiting from this bonus. There could get 5-30 (or more) people intentionally waiting and letting those people climb to be better targets, but there are other people who aren't in the loop who will still happen upon that Q and benefit. If 20 players get the benefit of that +10 additional points there are now +200 additional points in the slice. And each one of those 20 players is now worth more to anyone who might hit them. The economy of points in each time slice really does demonstrate trickle-down economics much more so than the US economy ever has.

That subtle improvement in the value of all your targets as an event goes on is helped by front-runners farming nodes to grow scores, and by them continuing to unshield to grow points later. I'm a believer in the fact that these efforts help pay a dividend to everyone else in the slice because that expansion of the point economy in the slice makes it easier for everyone else to reach a little higher, and maybe get to a progression reward that would have otherwise been difficult for them to manage.

woopie

VizMantis wrote:

fmftint wrote:

Grill reset

So you're saying he had a weak team of defenders out for other players to take advantage of and went against my low-risk team to get his strong defenders back out? That'd make more sense than iso farming I think. Although I don't totally understand what all grilling is. I understand it's the baking replacement, but somehow only benefits the super-high level players?

New system doesn't allow for a full cake, but it will allow for loaner char + 2 or 3* leveled char + champed 5*.

VizMantis

Dang Faustian, that's some extremely detailed analysis. It makes sense to me, if only it was as effective at inflating the point pool as baking once was. I never actively took part, but I miss the days when 900 points wasn't such a contentious score to reach.

Phumade

FaustianDeal wrote:

That said, I would disagree with the assertion that farming is only done "to kill time". The other reason why players will hit those teams is because they are trying to grow points in the shard.

Let's assume "farming ISO"
5* characters are incredibly demanding of ISO. Championing a 5* character has the same ISO cost as championing 5x 3* characters. It might not seem like a lot of ISO, but every victory counts, and the power differential between max champed 5s and 3*s or 4*s is so big that there was little actual risk - so it is productive in that regard.

Now let's assume "farming points"


That subtle improvement in the value of all your targets as an event goes on is helped by front-runners farming nodes to grow scores, and by them continuing to unshield to grow points later. I'm a believer in the fact that these efforts help pay a dividend to everyone else in the slice because that expansion of the point economy in the slice makes it easier for everyone else to reach a little higher, and maybe get to a progression reward that would have otherwise been difficult for them to manage.

farming iso: I relish the thought of maxing my next 5* 70 iso at a time hahahahahaa.

farming points: I don't disagree with any of your analysis, but I think your shoe horning a theory around the contours of the facts. The bottom line is do you think the 1200 player was sniping (supress,vendetaa, blah blah), Farming iso, farming points, doing a grill/reset, or another possibility against a 100 floater???? As a 5* roster, I've done/experimented with all of the above, and I would say sniping and farming can be long tedious work if you want the end result to be effective (and you know what I mean when I say EFFECTIVE). Of course, there can be alternatives and I'd be curious if you thought there was another explanation.


But I do think you raised an important point that should be discussed further.

Specifically, how sniping above 1k destroys points in the shard. I think that analysis alone is worth bring to all the check room mods as evidence why we need to reorganize around enforcer teams. Its no secret which shards have the highest points and coincidentally they have the most consistent enforcer players.

[edit] I've given this section more thought, and I want to quickly summarize your thoughts to make sure I understand the gist. 1. below 1k points, were actually trading the points and making a net gain to the shard.
2. Over 1k points, we are basically trading point points and its zero sum game.
3. Its the act of point dumping that actually causes the downward points pressure.

So in a sense the snipers are relying on the bakers unwillingness to dump points that they've baked in order to retal.

My tactical response to that would be to have the enforcer teams dump their points at much lower thresholds. While it may not be possible to eliminate the downward scoring pressure, it should be possible to mitigate the effects, and it would have the side benefit of maximizing the damage. In much the same way, people float at 600-800 to "take advantage" of some mmr quirks. Enforcers could split up by * tiers and focus on clogging certain point thresholds.

[end edit]

[edit 2] One positive thing we can accomplish is begin to refine the definitions of "growing", "trading", and "sniping" Specifically we can educate check room members to understand when some hits are actually beneficial to the room, when some hits are meant to inflict personal damage, and when some hits are really just dictated by the 'veil of ignorance'. I've often thought that the only difference between a panda and a killer whale is that the panda hired a messaging consultant. In the same vein, trap cake farms have existed since the dawn of alliances, but weren't a real thing until the community developed an easy to understand rational.

[end edit 2]


There is definitely rich iso and placement awards to be mined in understanding how player tactics affect the matchmaking and the points economy. You present a good way to differentiate between various forms of trading/sniping and it would help many players from low level 1* to the 500 level 5* adolescent whales that not all hits were meant with evil intent.

And that infact some hits actually grow the points in the shard.

FaustianDeal

There is an interesting semantic degeneration that happens in most discussions about the game. Players and alliances get reputations for sniping. But in some instances that is another example of oversimplification of language. But in truth, very few 'hits' are 'snipes'. *VERY* few. If you get hit in a PVP most of the time you were either:
* unshielded too long, with too soft a team, with too many points
or
* you just picked the wrong time to hop

Its not a criticism of anyone's roster. playing with 2x 450+ 5*s might deter some people from attacking you, but if you are worth 75 points to them you might suddenly be worth the risk. Just the way it goes.

mpqr7

I just lost 0 points to someone in PVP Sim.

I call it a "snipe" when someone hits me on a quick shield hop, but I realize it's not that they were waiting for me specifically to unshield. It's just that when I reach a certain point in the event, I'm at the top 20, so I'm more highly visible than others.. I might be one of the few people unshielded with that many points. It's just very crowded at the top.. unfortunately the rest of the crowd has much better defensive teams than mine, so I'm basically a cc at high scores.

FaustianDeal

Once you bubble into the top 20 and start taking hits from people in your own bracket, that's often referred to as a 'bracket snipe'. Its interesting, because there are a lot of players who have reputations for being bracket snipers.. but if they are not participating in any of the forms of outside communication and collaboration then they may just have (correctly) realized that watching their own leaderboard might be one of their best means of knowing when a good target is available.

(Unfortunately this is also the name applied to the practice of waiting for a fresh PVE bracket to open very late in an event and trying to secure high placement prizes for minimal effort.)