cp_gravelpit has been a popular map after the stopwatch was implemented, especially between high tier clans. One round consists of two parts: One team attacks and one defends, and then they switch it around. The question is whether it is best to attack first or defend first?
Some people argue that there is a psychological/tactical difference. If you were first to defend and the other team capped C in 4 minutes then you would play differently than if they capped C in 10 minutes. But is there an actual asymmetry in the implementation of the stopwatch? Is it possible that the stopwatch will determine a winner solely based on who attacked first?
To examine this I played around on gravelpit, trying out various scenarios. I came to conclude that the following statements are true regardless of whether you attack or defend first:
- If you capture more control points than the other team, you win.
- If both teams capture the same amount of control points, then the team that capture the “last” control point fastest wins. So if both teams fail at capping B, then the team that capped A fastest will win.
These two facts mean that in practice it doesn’t matter if you attack or defend first.
However the system does have a flaw. In the case where none of the teams manage to cap a single point then the team that attacked first will win. So you actually have a small advantage if you attack first but on gravelpit that rarely comes into play because the defenders almost always have to give up either A or B.
As mentioned, this probably never happens on gravelpit but it is not unlikely to happen on a map like cp_gorge. So if both teams manage to defend the first point on gorge then the team that attacked first wins. There is no logic to this at all. The only fair result would be a draw.
When we play best-of-2 in leagues we use the ABBA rule (so both teams get to attack first) hence none of the teams get the advantage described above. But in cups, when tieing is not allowed, the team that gets to attack first two times actually has an advantage. Now you know.
44 comments
TweetSolution = get rid of gorge until this is fixed.
- # - nice! +0+1.
Maybe they could transform the blue base till the bridge into a proper 5cp map
- # - nice! +0Umm I don’t think any league or cup would rule a 0-0 score as a win for either team. Or am I missing something here?
- # - nice! +0mikka: You are missing the effect of the stopwatch. On gravelpit if both teams only manage to cap A, then the stopwatch decides that the team that capped A fastest wins, and they will lead 1-0. On gorge if no one caps the first point, then the stopwatch decides that the team that attacked first wins, and they will lead 1-0.
- # - nice! +01-0 scoreboard wise sure, but why can’t people just use common sense and rule it as a draw… its not that complicated.
- # - nice! +0Oh I didn’t realize it puts a capture point on the scoreboard. That’s a weird bug.
But yeah this really doesn’t affect any actual results – just the screenshot.
- # - nice! +0i think forcing people to play gorge should be reconsidered for next season at all
with an eye on esea you can see that there is a discussion about putting mainline into the mappool. also coldfront and prolane are two good maps which would fit a lot more into etf2l mappool than gorge does
- # - nice! +0This issue hasn’t been addressed before. If you look in the ETF2L rules it says: “cp_gravelpit and cp_gorge have to be played with 2 rounds (possible overall results: 2:0, 1:1, 0:2).”
If it was possible to draw rounds then results like 1:0 or 0:0 would also be possible. So clearly they haven’t taken this into account.
- # - nice! +0Interesting. Now that it’s known, i don’t think it will be a practical problem, as admins should be able to understand it.
The biggest problem with attack/defence maps is that the results are basically a dice roll. Lets say statistically that you win every 3rd attempt at an attack. If the statistics layout an equal amount each number pr attack attempts, that means that your team’s worst scenario of attempts = 5.
So, completely based on random luck, you could have a round of doing 5 attempts on both caps – or you could be in luck, and take it on 1st attempt both times. -Dramatically changing the times.
It’s a shame that there is a system for such randomness even at the highest divisions. It would be like every 5th premier league match – they’d play with different rules and 4 goals.
- # - nice! +0Yeah and in 5cp maps 3-2 scores etc are also based on random luck. Why do we bother? Let’s just throw dice…
- # - nice! +0There’s a clear difference between being “based on luck”, and having “an involvement of luck”. My parallel to football should illustrate that perfectly well.
- # - nice! +0I disagree with that statement Torden :) Even though it’s based on attempts, so is every single other normal 5 cp map. You load uc, defense is in place, and you push in, sometimes works, sometimes not. Only difference is, on 5 cp enemy can push back and recap if your attempt fails.
But it’s still down to being able to doing a nice rush and get things done. Sure there’s an involvement of luck, but so is there in normal maps. There always is :)
Just my 2c
- # - nice! +0Sure. But, consider the fact that attack/defence maps are solely won on time. Time on 5cp maps are only one of several factors. Also, consider the statistics of “weaker” teams managing to steal points from “stronger” teams on it.
This season alone has some upsets: Blame stealing from Rockit, nWave stealing from PENTIA (arguably), C2 stealing from Broder (that’s 3/4 of div2 games played that week). If you continue looking down the divisions, you’ll find a trend in results on Gravel Pit not matching the results of the 5cp map.
It’s ok to be biased based on what map one likes or doesn’t like. But guess what – you can’t really argue with statistics.
- # - nice! +0AvD maps aren’t based on luck any more than 5 point CP maps. The difference as has been pointed out is that the opportunity to correct a mistake is almost non-existent on an AvD map. Does this make them any more based on luck? No. The opponent is still beating you with skill, they didn’t randomly kill your medic with a spy, they did it on purpose and by using tactics.
As for upsets being more likely on these maps, again that’s not a bad thing. A good team is a good team on all maps. AvD maps create extremely intense and high pressure situations where a single mistake will cost your team the round. Isn’t that the best judge of ability, rather than assuming that because a team is strong on 5 point CP maps that they are the ‘better’ team?
If there are so many upsets, perhaps clans should take it as a sign to practice these maps more.
- # - nice! +0“AvD maps aren’t based on luck any more than 5 point CP maps. The difference as has been pointed out is that the opportunity to correct a mistake is almost non-existent on an AvD map. Does this make them any more based on luck? No.”
That’s an interesting debate technique, where you list a number of statements, then give an answer based on the statements being true. But fact remains, you still don’t give any new actual arguements – just arguable statements.
” A good team is a good team on all maps.” -A fair statement, but bear in mind that you also claim the same thing to be true for Goldrush, Dustbowl, not to mention the vast amount of rarities out there.
And don’t forget that the statement you’re arguing against isn’t that “good teams shouldn’t be good on all maps” – but that there are certain maps (like Gorge and GP) that involve more element of luck than current 5cp maps. Thus, your discussion of what-maps-a-team-should-be-good-at becomes irrelevant, an interesting off branch at most.
I don’t deny the chance of my statement being wrong, but it would be highly appreciated to see someone coming up with facts or numbers to support their counter statements.
- # - nice! +0“they didn’t randomly kill your medic with a spy, they did it on purpose and by using tactics”
Surely a big amount of luck is involved too. If a player was able to kill the medic as spy 50% of his attempts on gravelpit, I would recruit him instantly! It does take lots of luck.
I have to agree with torden that the outcome of gravelpit heavily is dependent on luck – getting a lucky headshot, backstab or crit sticky. However I wouldn’t want gravelpit removed. Due to the variety of maps, div1 has never been more exciting than now.
- # - nice! +0Yes, i agree fully with that. Not so much from a player’s point of view, but as a spectator.
- # - nice! +0Luck is success arrived at by chance. To use F2′s example there is no more chance in a crit sticky, headshot or backstab happening on cp_gravelpit than there is of it happening on any 5 point cp map. The only likely difference between the two is that on a 5 point cp map it would cost them a round, on gravelpit it costs them the chance of doing any better than drawing the map.
To use an example, JH is an awesome sniper. You can’t say he’s a lucky sniper, he just plays the class very well. Why then is his skill considered a random success on gravelpit but not on badlands?
- # - nice! +0so, you wanna use crits too?
- # - nice! +0The point I’m trying to make is that too many people (including myself and my clan in the past) labour under the pretence that they can only do so much to help their team win gravelpit, the rest is up to ‘fate’. This isn’t true, it’s just defeatism.
Poker is a game of chance; with the exception of cheating there is no way you can influence the hand you are dealt and must play with. TF2 is a game of skill, you can practice, you can adapt to snipers, spies and kritz. Sure, on gravelpit you have less time to correct mistakes, but doesn’t that make it all the more important to make sure you get it right the first time?
The random number generator doesn’t beat you, your opponents do.
- # - nice! +0That’s not my point torden, but it is a perfect example of luck in games. Enabling random kritz would re-introduce a great element of luck into TF2 where your success is dependent on a random number generating function inside the program, playing gravelpit isn’t the same.
- # - nice! +0did you read this?
“The biggest problem with attack/defence maps is that the results are basically a dice roll. Lets say statistically that you win every 3rd attempt at an attack. If the statistics layout an equal amount each number pr attack attempts, that means that your team’s worst scenario of attempts = 5.
So, completely based on random luck, you could have a round of doing 5 attempts on both caps – or you could be in luck, and take it on 1st attempt both times. -Dramatically changing the times.”
- # - nice! +0The huge difference is that if your medic gets backstabbed on gravelpit you will most probably not be able to win the map. If your medic gets backstabbed on granary, you might only lose 1-2 cap points, and not even a full capture.
The bottom line of my point is: If granary and badlands were played like 2 golden caps, so the only possible results were 2-0, 1-1 or 0-2, then there would be a lot more luck involved in winning. If gravelpit were played best-of-7 then the best team would win more often.
- # - nice! +0Sometimes, the numbers really do tell the true story.
- # - nice! +0‘sometimes’? torden? i think ull find the stats NEVER lie :D
- # - nice! +0“The biggest problem with attack/defence maps is that the results are basically a dice roll. Lets say statistically that you win every 3rd attempt at an attack. If the statistics layout an equal amount each number pr attack attempts, that means that your team’s worst scenario of attempts = 5.
So, completely based on random luck, you could have a round of doing 5 attempts on both caps – or you could be in luck, and take it on 1st attempt both times. -Dramatically changing the times.”
Torden. You are using interesting debating techniques too! Reasoning in circles. Results are bases on luck, so we can aply statistics! Hey statistics means something with luck! So luck is involved! Hey thats what i wanted to conclude!
Come with better arguments, because you arent proving anything. I can make the same arguments against you so its a bit pointless to throw that argument into the discussion.
I have a much better explanation. I think 90% of the teams dont like gravelpit. Its just a wild guess. They dont like to practice it and they are only playing granlands and other 5-cp maps. So when they need to play gravelpit, they practice it maybe one week. Maybe not at all. At least not enough to get the hang of the map. With the timer, the very big layout and the attack/defense principle its substantially different. So both teams dont know what tactics work and what tactics dont.
For example. One team seems to have a very good defense, while the other team cant find the tactic to counter them. While another time, one team just throw some random defense which sucked ass and the other team seems to have a tactic that seems to counter the defense and setting a very fast time. Everybody are doing some random tactics and thats what you can call luck. But still the random decisions of the teams make them succeed or fail. In the end, not different to other maps.
- # - nice! +0The problem with gravelpit and other stopwatch maps is that there is no good way to recover from a screw up. If you screw up on badlands mid fight you can still contest spire, and then contest your last point. Even if you lose that you still have another round you can win (and you can even end up winning the map). On gravelpit its basically over or at least a draw when one mistake is made.
- # - nice! +0“Torden. You are using interesting debating techniques too! Reasoning in circles.”
“I have a much better explanation. I think 90% of the teams…”
You’re countering numbers with “i think”. Cute :)
- # - nice! +0So to come to a conclusion… playin AvD maps is like playing 2 fullcaps on any 5cp map.
Scores can be 2-0, 1-1 and 0-2
- # - nice! +0“did you read this?”
I’m afraid I’m not sure I understand what you’re saying with that example, could you explain it in a different way?
If the fairness of a map (the best team always winning) is judged by the chance to come back after losing points, then is ctf_turbine is the most fair competitive map?
- # - nice! +0In a way it is. Top level matches are often quite even in scores and comebacks happen.
Gravelpit is often decided on a critical mistake or two, while 5cp/ctf modes give you much more time to correct them. I’m not saying gp is bad, I love it, but it is a lot more “random” than other maps.
- # - nice! +0Exactly Sturmis
- # - nice! +0gaylord
- # - nice! +0Gravelpit is often decided on a critical mistake or two.
And why is this random? Isn’t the teams fault for making those mistakes?
- # - nice! +0The scores on gp might be random but the mistakes are not.
Last I checked it was possible to put any old score you wanted in for attack defence maps on etf2l. So as long as the teams are reasonable (a big ask for a lot of people on the internet I admit) and agree that double fail is a tie not a victory for the first attacker this entire point is moot. An explanation can easily be put in the match comment section explaining what happened, that’s what it’s there for after-all
Also torden nice side stepping on Waster pointing out your entire argument is based a “lets say” Mocking his argument for being as baseless as yours isn’t very good debating skill either when you do it that obviously, regardless of what politicians may think. :P
Teams don’t cap any point on GP every third time roughly averaging out to 5 for B and C with A never defended. They attack each point until they cap it, running through their strategies till one works or the clock runs out. That’s where the luck is, will you get the successful strategy first time or the 9001st or never. But even then the better team will see the weakness in their opponents defence quicker or be using better strategies to begin with so it’s still not truly random or luck based.
Kills a player gets are result of skill on their part (or occasionally a lack of it by their opponents) luck isn’t involved, we’ve spent years pestering valve for settings to make it that way.
- # - nice! +0@Krakyn:
“Also torden nice side stepping on Waster pointing out your entire argument is based a “lets say” Mocking his argument for being as baseless as yours isn’t very good debating skill either when you do it that obviously, regardless of what politicians may think. ”
Hehe, it’s turning into an interesting meta discussion this. What’s ironic though, is that while those discussions actually have some relevance to the question at hand, you go straight after trying to call who uses the worst debating technique for putting fourth one’s arguement.
-You can agree or disagree with those techniques, but it does not change the arguements nor the numbers.
you continue by saying this:” That’s where the luck is, will you get the successful strategy first time or the 9001st or never. But even then the better team will see the weakness in their opponents defence quicker or be using better strategies to begin with so it’s still not truly random or luck based.”
Ok, so first, you acknowledge the fact that there is an element of luck involved. Then, to “win” the arguement, you put in your own conclusion “so it’s still not truly random or luck based.” -Which will trump my first statement of (“The biggest problem with attack/defence maps is that the results are basically a dice roll.”)
Fair play. The first statement of it being a complete dice roll, is in fact exaggerated. Good to have that sorted once and for all.
But additionally, if one were to interpret what you write correctly, you do say that A/D maps have some element of chance, but good teams will be better at finding the better way.
But, because its not the caps, but indeed the timer that sets the bottom line premises, that variable of chance may not reach it’s destination before the bell rings. Hence, the element of chance is larger than that of a 5cp map.
Why make such a large discussion on such a small detail, i don’t know. But, it sure is entertaining as hell. Peace! :D
- # - nice! +0btw, F2 formulated it brilliantly quite a few posts back:
“The bottom line of my point is: If granary and badlands were played like 2 golden caps, so the only possible results were 2-0, 1-1 or 0-2, then there would be a lot more luck involved in winning. If gravelpit were played best-of-7 then the best team would win more often.”
Instead of trying to go after who said what when, I think the fruit of the discussion reached it’s peak back there. So, if any of you wants to keep taking my baits (or try to bait me back), i’ll be more happy to reply – Just bear in mind that so far the wisest thing seems already to have been said.
- # - nice! +0““Torden. You are using interesting debating techniques too! Reasoning in circles.”
“I have a much better explanation. I think 90% of the teams…”
You’re countering numbers with “i think”. Cute :) ”
And again. Maybe you should search the web for fallacy. Because thats what you are doing the whole time now. Manipulating and executing incorrect argumentation to make yourself look good. You have given one argument which i cracked into a fallacy and after that you didnt give any new one.
More over, you have clearly not read my post. At least i am honest about what i know and what not. Should i ask from 500 teams registered in etf2l if they like gravelpit or not? Thats why i made an assumption there. But it totally isnt my point. You can just ignore that line, because my point still stands. No need to attack someone personally on a line thats irrelevant.
- # - nice! +0A’ight, so. I have zero interest in kiddies throwing ad hominems and non sequiturs around like candy, or your metadiscussion.
snowiE argued that luck is no greater factor in AvD than in 5cp, which is obviously incorrect given its vast difference in magnitude and influence on a result. Even if I were to accept the assumption snowie makes that the base chance is unchanged (but map flow [by this I mean that there are lots of waiting periods for the waves of attacks, and the attackers can't defend pre-emptively by pushing back, such as on 5 cp maps] and psychological factors will in all likelihood affect this), the impact on a game to have your medic die once to this 0.2*0.5 chance in AvP is far greater than in a 5cp, and thus by this property it follows that luck plays a more significant role in AvD. This is because branches of events with a certain chance grows from every decision, and they occur in every game, only they play a much larger role in AvD.
One lucky sticky hit (not even crit sticky) can assure a team of a draw, regardless of what happens the next 20 minutes. For 5cp maps such is not the case, so obviously it is not as affected by elements of chance.
- # - nice! +0Oh, and, the 0.2*0.5 was supposed to be 0.X*0.Y just to illustrate that there are several variables in play, not to give concrete data.
- # - nice! +0dem be trollin
- # - nice! +0Waster, you say this:
“Maybe you should search the web for fallacy. Because thats what you are doing the whole time now. Manipulating and executing incorrect argumentation to make yourself look good.
then, magnifically, you end by this:
“No need to attack someone personally…”
Well played. I’m really rooting for your next try, though. you know what they say – 3 times = the charm!
- # - nice! +0Oh, and Waster
“Should i ask from 500 teams registered in etf2l if they like gravelpit or not?”
I’m guessing that is you being sarcastic. But, in fact there’s no need to be so negative. If you want to learn that info (or get an indicator), all you could do is to ask me or ETF2L to make a poll on it @ respective sites.
So, with the absence of you asking, my answer is “yes” – this could be an interesting poll after the current one. The world is such a beautiful place when people work collaboratory, no? :—)
- # - nice! +0