Does ELO mean different skill level in different games?

[cranesflyhigh]

Okay, so here is something I have noticed playing multiple strategy games on BGA (given I am not the best in any, but I am somewhat okay I'd say). In Terraforming Mars people at around 200 ELO still make really attrocious mistakes. Like spam cities with SP at the start of the game or draft and play Hackers. In another game, say Agricola, a comparable mistake would be to spam start player without a significant reason or play way too many questionable quality occupations early that hinder other development. But 200 ELO players in Agricola seem to understand these concepts, these are more relegated to 100ish players.

My hypothesis is that ELO scales differently according to player base. Terraforming Mars while still new, has a massive player base. Agricola, while a really great game, sadly isn't played as much. So the distribution of players and their skill level is spread way wider in a large player base game than in a smaller one.

Is my analysis and hypothesis correct or am I missing something?

[Manolo de Xaninho]

I must confess myself as a Hackers supporter (sometimes). I know the card is highly controversial,l but there are situations when is perfectly viable, imho. There are more exciting openings for sure, but if you have nothing better to do at the start of the game, i think it can be an option. I would not judge the level of a player precisely by the use of that card.

Answering your question, i think the level of players of a certain ELO (not limited to 200, but in general in this site) can be highly variable. And yeah, most people here play too many cities, too often, thats probably the major mistake a lot of people do make here. And is easily flixable. People, SP cities is almost always a mistake, barring niche situations or in the last round with nothing better to do.

[cranesflyhigh]

Hackers for sure is playable in very rare cases, where the stars align - spare power prod, probably have drawn the card instead of draft, can afford the -1point and the 4mc prod swing actually matters. That is a lot of ifs, and is why the card is generally awful, but even awful cards have uses sometimes. What I was mostly referencing is playing highly prioritizing Hackers, drafting it and actually playing power prod before just to get it down. Which I do see a fair bit from 200ish players and I do consider a really bad play since it hurts the user most likely way more than it benefits them.

The variability I have encountered too, but that is common across every game I think. I was just thinking about how the variability is very different across games with different player pools. I dunno, just a theory.

[Yahyra]

Hate to be that person / beat a dead horse... but... player count matters. Hackers is a (in my opinion) perfectly playable (early/mid) card to play... in 2p TM. Any higher and of course, the card is terrible. Most attack cards are much, much better in 2p. etc etc. .

I do think the ELO layers will disseminate as you said Crane. Another game I enjoy a lot is Lost Ruins of Arnak. A year ago players even at 400 ELO were not that strong. Now everyone is much stronger, make less 'fundamental' mistakes.

Keep on with the awesome discussion. Best wishes.

[cranesflyhigh]

Player count matters for sure, in 2p Hackers rises from being an awful card to just pretty terrible. The tempo on it is rough. 13mc (3draft cost + 3 pay cost + 7 for the power prod) and -1point for a 4mc prod swing is just bad value, period.

Yeah, it also might that TfM is just new, so the elo is very wonky currently, good point!

[Yahyra]

Context is everything... in a 2p game, Hackers being a 4 MC swing... from generation 1 in a longer game? Perfectly playable card.

In a 2p game, Hackers flipping a funded banker award? Perfectly playable.

In a vacuum without context, in 2p... it can be ok / playable. Many other "playable scenarios" (got OP science cards, +9 energy prod and not much to spend it on? also makes Hackers more playable, etc. etc.). Best wishes.

[charliec247]

Okay, so here is something I have noticed playing multiple strategy games on BGA (given I am not the best in any, but I am somewhat okay I'd say). In Terraforming Mars people at around 200 ELO still make really attrocious mistakes. Like spam cities with SP at the start of the game or draft and play Hackers. In another game, say Agricola, a comparable mistake would be to spam start player without a significant reason or play way too many questionable quality occupations early that hinder other development. But 200 ELO players in Agricola seem to understand these concepts, these are more relegated to 100ish players.

My hypothesis is that ELO scales differently according to player base. Terraforming Mars while still new, has a massive player base. Agricola, while a really great game, sadly isn't played as much. So the distribution of players and their skill level is spread way wider in a large player base game than in a smaller one.

Is my analysis and hypothesis correct or am I missing something?

I agree with your analysis. The more players there are, the more opportunities to get wins. I usually look at what the ELO of the top players are to determine the expected skill level for other ranges.

In terms of expected skill level, you'd probably have to factor in whether the opponent was experimenting or otherwise playing sloppily for other reasons. For example, I play really sloppy when I am on my iPhone because the screen is too small and I end up not bothering to check every detail of the game state, instead of choosing to go with the flow; or I am playing while taking public transit and lose track of a detail or two when I have to get on/off transit and those end up being game-changing. On the flip side there are games where I am totally focused or locked in, doing all the math and taking notes.

But I agree that at a certain skill level there are some moves, such as SP City, that you almost never make.

[EvrikSan]

ELO mean mostly nothing cause you can always "farm" it while playing with your weak friends

[BlackHatGuy]

In a word: Yes.

ELO is a purely *relative* measure, that only applies within whichever pool was used to generate it.

A 400 ELO player should, on average, win about 2/3 of their games against a 300 ELO player. And be beaten 2:1 in turn by someone with 500 ELO.
But 400 itself can mean world-class, skilled, average, or a rank beginner depending on which game it's being applied to.

The way to think of it is like this:

The range of an ELO pool (IE how high do the numbers go) is a measure of how finely you can distinguish the skill levels of different players.

More players = higher ELOs because it gives you more data to work with.
With 100,000 active players you can identify a top 1%, 0.1% or even top 0.01% player.
If there were only 100 players, you probably couldn't go beyond top 5%.

So with 100 players, ELOs might not get beyond the 300s. But the same game with 100,000 players might go into the 600s or more.

The other main determinant is how skillful the game is. IE how much of it is luck.

If there is 0% luck, then a win gives you absolute certainty about who played better in that game (see: chess, go, etc.)
If there is 100% luck (say, flipping a coin), then a win gives you zero information about who is the better player.

Despite an absolutely miniscule pool of players, chess on BGA still has an ELO range at least into the 400s (based on players I could track down).
Whereas Catan, with a much larger active player base, barely gets to 400 in the first place.

So to go back to your original question:

Yes. 400 ELO (for example) tells you *nothing* about how good someone is at a game (except that they're not the absolute worst at it). You have to know what the ELO range for that game is before you can know what the 400 means.

[FrostTomcat]

Hacker supporter +1 lol!

While I never play SP energy and then hacker, I believe in the game with extra energy, it is highly playable.
In that case, it actually turns one heat production into 4 MC productions while cost 1VP, and in a long game, this 1VP can be easily regained after build my engine.

Some other times If I got hackers + two or three non-critical cards, I may keep the hackers first (but not playing it), and leave the remaining cards to my opponent to choose. Being hacked in the early stage without good ways to gain terraforming rate and / or MC production can be devastating.