People are literally arguing against mathematics in this thread. All things considering, it is very impressive.
Go back to school, guys.
Go back to school, guys.
POLL: Which team is better-off after a 0-0 in the first leg of a two-leg tie?
- Thread starter BaseFishing
- Start date
-
In the latest episode of the United Hour Podcast, Nik @Rood and Imran @Annihilate Now! start with the good news by celebrating Manchester United Women's emphatic FA Cup win over Spurs at Wembley and the future of manager Marc Skinner.
Then it's the bad news and the ever worsening form of our Men with losses to Arsenal and Palace.
Finally the main debate - ETH Out now or keep him until the end of the season - if not longer. Are you Team Nik or Team Imran?
I meant the away team in the first leg, so I agree with you
On average, no.
You decided so, and then pulled data out of nothing to confirm your prior belief. It is the definition of 'if facts don't suit your theory' change the facts'.
Every data that has been put here, with and without last 16 (which was your original argument, until you were proven totally wrong and then changed the argument), confirms that the home team in the second match has an advantage. Removing last 16, just dropped the advantage from 67% to 65%. Someone even mentioned a journal paper which did the same experiment including counting only 'evenly matched teams' (whatever the feck it means), and still, surprise surprise, home team in the second match has an advantage.
I love the way "we have explained this multiple times now" is supposed to convince you. As though saying the same wrong thing repeatedly makes you less wrong.
Teams at home tend to play far better. I'd rather be in Barca's position right now than Liverpool's.
Bayern and Barca are expected to win at home, Liverpool and Lyon are not. Most teams have different forms when away vs at home.
Bayern and Barca are expected to win at home, Liverpool and Lyon are not. Most teams have different forms when away vs at home.
I have actually been thinking about this a lot today, as there were a couple of things I could not reconcile, despite being comfortable with the fact that a 0-0 in the first leg will on average see the home side in the second leg progress 67% of the time and the away side 33% of the time. The historical data shows that to be true, even when removing seeded ties where the better team players away first from the equation. I think the question is still why though? Does home advantage so comfortably outweigh score draws being a defeat? I'm not sure.
I think Pagh Wraith could have a point. I think the data could be corrupted in the sense 0-0s happen more frequently in the first leg when the weaker team is at home and the stronger team is away, regardless of whether they were seeded this way. I think that Pagh could also be barking up the wrong tree, so I looked elsewhere at the pattern of home wins, score draws and away wins in football matches in general.
What I've not be able to explain is why does the home team in the second leg progress two-thirds of the time? Do home teams generally win two-thirds of their matches, with score draws and away wins making up the other third? Not really. I've not got CL figures, but in the PL in 17/18 46% matches were home wins, 18% score draws and 28% away wins. That's 46% vs 46%, with scoreless draws being split for the sake of argument. La Liga in 17/18 was 47% home wins, 16% score draws and 30% away wins. That's 47% vs 46%. At random I looked at Serie A in 12/13 where you get 47% home wins, 16% score draws and 28% away wins. That's 47% vs 44%. Clearly home wins vs score draws and away wins are balanced pretty evenly. Why therefore are we seeing the home team in a second leg progress after a 0-0 away at a much higher percentage at 67% or so? The pattern of results in normal football, albeit league matches, does not support this.
I'm wondering if the distribution of match results between top sides, i.e. sides that are more closely matched than those on average in a top European league, which is the Champions League knockout rounds, is quite different. In matches between top sides, in the league, in a cup, in Europe, wherever, actually looks more like 60% home win, 12% score draw, 18% away win, so 60% vs 30%, with the goalless draws split. After all, the number of away wins you would expect to be less when you removes fixtures from a sample, like United and City going away to Cardiff and Huddersfield, but instead only away to Liverpool and Chelsea. Here you are focusing on fixtures where the home advantage is the bigger difference between the two sides than actual footballing ability. On a neutral ground, the sides would be close to 50/50. In a home/away match, that is very different.
My hypothesis would therefore be that sides progress 66% of the time at home in the second leg after a 0-0 draw away in the first leg, because European two-legged matches tend to feature much more closely matched sides, than you average league match where sides are more disperse in ability, and where the pattern of home wins vs score draws and away wins is quite different
Regardless, this doesn't change the fact that if Team A vs Team B finishes 0-0 in the first leg, you should take Team B to go through when they play at home in the second leg every time, if all things are otherwise equal and you do not know anything more specific.
Unless your name is reddevilboy in which case Bet365 will allow a 50k stake and automatically upgrade you to a VIP account.
Strange on so many levels
Very nice. Thanks for taking the time. Who did you consider top teams when looking st the results?
This is indeed what I am saying. And there is no doubt about it. The Liverpool v Bayern tie is the best example.
Not completely bonkers. On the second leg, the first leg home team (FLHT) only have to play to a scoring draw to advance without PK's. The first leg away team (FLAT) has to win at home to advance without PK's. To offset that, the FLAT has home-crowd advantage on the second leg. All else being equal, FLHT has two ways to advance, while FLAT has only one. But...
I'm in the "impossible to say" camp, because my above argument depends on all other factors being equal. They never are.
That isn't how Pinnacle do it though, they use their sharp players to adjust their lines. I think you could make a strong case that their method is optimum, if you could be arsed like.
GUYS.
There's an actual study done on this by the "European Journal of Operational Research" that included ALL European ties since 1965 (when the away goal rule was introduced.) It details different likelihoods based on the quality of teams in question. For example, it is subclassified into "Teams of equal strength" and "Weak home team & Strong away team". For teams of approximately equal strength, the likelihood of the home team that attained the 0-0 draw in the first leg (in this case, Liverpool) progressing is approximately 45%.
So basically Bayern have a very small advantage if you think Bayern and Liverpool are of equal strength.
There's an actual study done on this by the "European Journal of Operational Research" that included ALL European ties since 1965 (when the away goal rule was introduced.) It details different likelihoods based on the quality of teams in question. For example, it is subclassified into "Teams of equal strength" and "Weak home team & Strong away team". For teams of approximately equal strength, the likelihood of the home team that attained the 0-0 draw in the first leg (in this case, Liverpool) progressing is approximately 45%.
So basically Bayern have a very small advantage if you think Bayern and Liverpool are of equal strength.
None of this post makes any sense. If I understand you correctly, you think that Liverpool are the better side and were the favourites before kickoff. The betting markets disagree. The first leg was priced as an evenly matched affair for a long time (meaning the home team is around 2.20 to win) with a very late drift in Liverpool's odds. They closed at 2.33 which means Bayern were rated as the slightly better team at kick-off. If this is true, then 0-0 is a slightly better result for Liverpool which could explain the slight shortening of Liverpool's odds to progress. This is so marginal though and within the margin of error that it barely matters. For the purpose of this thread these are two equal teams and the first leg result did not alter their chances of advancing to the next round. If we look at the second leg, we see that Bayern are 2.10 (implied chance of ~47.5%) to win (so again they are rated the better side by a fraction). And if we add up all the different scorelines that see them progress (~47.5% wins + a little more than half of the 0-0 results) we arrive at the same 50% we had before the ball was kicked. I don't know how many extra-times and penalties are won by the home side, but even if it were 100% that number would only rise to ~54%. The first leg result had NO effect on the teams' chances. If you go through historical odds and results of other evenly matched ties that were goalless after the first leg you will find the exact same thing. I'm repeating myself here but the bookies are not wrong. There is a direct 1:1 correlation between closing odds in high-liquidity markets and actual results.
The reason that an average of 65% of away teams go through after a 0-0 away is that these teams were expected to go through 65% on average in the first place. Because of the away goal rule 0-0 is the only result that has no substantial effect. All other scorelines do. And if the away goal rule didn't exist 0-0 would be a brilliant result for the away team of the first leg and their expectancy would go up from 50% to roughly 60%.
This is getting super tiresome and I'm done for now here. I'll check back tomorrow to see if there's any progress been made in this thread.
Last edited:
We do need a lot more examples though. There are other factors effecting the odds in a one off match.
Your argument started off as the seeding contaminated the sample.This is getting super tiresome and I'm done for now here. I'll check back tomorrow to see if there's any progress been made in this thread.
Then it changed the fact that stronger teams are more likely to get 0-0 away.
If that wasn't the case then nobody understood that and you failed to correct them when they didn't.
Both valid arguments, don't get me wrong, but the confusion in this thread is at least partly to do with you and your inability to articulate yourself well, and not just, as you're implying here, that we're all a bit thick.
Correct, there are other factors in play and new information that wasn't present in the initial pricing, such as each team having a relative home/away strength, and the very fact the betting markets have SEEN both teams play each other once and one team might have performed better than expected, also the suspension for Kimmich would cause a relative drift for Bayern compared to what they would be with him.
We both thought the initial data discrepancy (the very fact the odds are 50/50 is enough to be sceptical about data that finds the chances being 67/33) could be attributed to the seeded draw in the last 16. When it became clear that this didn't really impact the data he came up with the idea that 0-0 draws are just generally more likely to occur with the favourite playing away from home.
Both arguments are correct as you said but yeah I can understand that the change in argument could cause confusion.
Yeah true, sports trading is a bit of a joke in the west these days, it's all completely automated to just follow pinnacle who do all the actual work.
IF we agree home advantage has declined gradually since 1965, that 45% chance probably increases a touch, if the approximated figure is in the 47-48% chance area, given error bounds, variance etc (defining teams of equal strength is the big error bound), is this really enough to say statistical significance that the real figure differs from 50/50?
Also on the point of teams being equal strength, surely this would need to be uniform across the matches to be fair? the fact Liverpool have Van Dijk back whilst Bayern lose Kimmich means the relative home/away strength will favour Liverpool for the next match more than in a normal 50/50 tie. Think that makes sense.
Fantastic post. Maybe the home advantage in European matches is a little bigger than in domestic leagues simply due to less familiarity with the environment for the away team?
Also, why does a question like this even need a poll when there is a clear answer?
It's so weird to me, people treat 0 - 0 as if it's a bad result for the away side whereas when they lose 2 - 1 that's almost always a great result. People really seem to have so much trouble grasping the pretty simple away goal rule.
Last edited:
Just be clear, I’ve not looked at the distribution of home wins, score draws and away wins between top teams. I was just guessing what it may look like. It would be labourious to do it manually, as I don’t have a data set for it. I can’t even find data for matches in the CL aggregated like that, only for domestic leagues.
It's genuinely fascinating to me how even the vote is here. Whether or not one answer or another can be shown to be mathematically correct (can't be arsed to trawl through the whole thread), it's interesting that in people's perceptions home advantage and score-draw advantage are so evenly balanced.
On average, I would say the home team in the first leg but it still depends on countless other factors. If the tie is between two teams with really strong defenses (Atletico-Juve tie for instance) the advantage is probably bigger for the home team because having the away goal advantage is crucial in a tie with few goals. In higher scoring ties, (PSG vs City as an example), I would give the edge more to the away team.
The strength of each team home vs away is also huge. Given our away record for example, I would probably take a 0-0 away against most top teams.
The strength of each team home vs away is also huge. Given our away record for example, I would probably take a 0-0 away against most top teams.
The team that draws 0-0 at home in the first leg should be in the better position as there are more types of favorable results in the 2nd leg that benefit them (draws with goals) but that's not the reality after you factor in traveling, the home ground, fans and how it affects player mentally coming into the match. Statistically the away team that draws the first leg usually go through. Of course things such as seedings also factor into that.
That's a really good question. It depends who you are playing. If you were away to a top club and drew the first game they could always score away against you leaving you chasing the game at home. Think 2000 when we played at Madrid and drew 0-0. They came to OT and won 3-0 I think it was. It's only a good result away if it's an easy team you are playing at home 2nd leg
I knew the data was wrong based on logic and math. So the influence of seeded teams was the obvious explanation. And it does have an influence just not to the extent I expected. So then I looked for another aspect that explains the discrepancy and found it. It is a complex issue and I've been trying to articulate myself well. It's just super frustrating to argue with people who are not as educated about a particular subject (and there is nothing wrong with that!) as yourself when you know you're right. We have argued with numbers and hard facts in between noise of "yeah but if the home team couldn't score against you on their home turf they are obviously weaker so 0-0 is good for the away team" (paraphrasing here) which is just random words really and doesn't advance the conversation at all. Then you get people trying to explain how bookmakers work (or that these markets don't mean anything) to others who actually work in the betting industry. I think I know a bit about betting markets and I wouldn't dare that.
We have proven it in any way possible now (see below for added evidence based on actual historical results). And I hope the people screaming the loudest at least realise they were a bit ignorant.
Pretty much this. Home advantage is much lower these days than it was in the 70s and 80s (see here: http://old.clubelo.com/Articles/Themysteriousdeclineofhomeadvantage.html). I used 0.35 goals for my calculations which is a good average in modern-day football and got a chance of progressing of 51% for the home side from the first leg. If I change the home advantage to 0.5 goals that number declines to 48%, if I use 0.7 goals, it goes down to 44%! Also I (conservatively) assumed that the home side wins about 55% of extra times and penalty shootouts. The number could be higher in reality moving the chances of both sides to ~52/48%.
Last edited:
We've been through this in the CL thread. It's not against mathematics. Statistics are a different animal. People are confusing statistics of actual outcomes with being favored by the possible results.
@Pagh Wraith the link is broken
I've seen you say it's proven beyond any doubt a few times and I'm inclined to believe you, but the thread is a mess and it's hard to discern when or where that happened, to be honest. ( I'm not asking you to go over it again, just saying I can see why people are confused )
I've seen you say it's proven beyond any doubt a few times and I'm inclined to believe you, but the thread is a mess and it's hard to discern when or where that happened, to be honest. ( I'm not asking you to go over it again, just saying I can see why people are confused )
Does anyone have a link to the paper?
Fixed
Is the away goals system actually that massively significant?
Do we have any interesting stats about how often it's actually used to settle things? (did anyone mention something earlier that I;m now missing?)
How many times have Utd lost out or proceeded through it?
The fear/dislike of it seems excessive to me. We do talk about it a lot though.
Do we have any interesting stats about how often it's actually used to settle things? (did anyone mention something earlier that I;m now missing?)
How many times have Utd lost out or proceeded through it?
The fear/dislike of it seems excessive to me. We do talk about it a lot though.
Depends on how you define significant.
Here's how it influences the chances depending on the first leg result. Again we are talking about two absolutely equal teams. I used Liverpool and Bayern to avoid any confusion about who is home and away. Stating the obvious here but based on the numbers the away side should really be incentivised to score in the first leg.
Last edited:
cheers!
0-0 at home means that the opposition is forced to win the tie. Draws with goals will result in a win, so getting the first goal means the other team need 2 to not be losing.
Hm?
You folks are valuing being at home too much.
Get a early goal against and you lose unless you score two. If they could hold you to 0-0 then the probably they can keep you from scoring 2.
Last edited:
How many of those draws are goalless?
I have them advancing 12 times at home vs 14 away from home after a 0-0 first leg draw, assuming none of the draws are goalless.
I don't rate home advantage as some do. Ultimately it comes down to the strength of both teams, but the disadvantage of having to win the game unlike the other team more than balances out home advantage imo.
Yeah. That is my view as well.
Pretty much.
No, he never proved anything, just kept saying he did. And keeps using the betting odds for Bayern v Liverpool as proof for some inexplicable reason. He even used it now when someone asked how significant the away goal system generally is. It's a bit bizarre.
That's an odd argument. Surely you can then say that if a team couldn't score at home, surely the home team in the second leg can keep them from scoring away.
It's not exactly guaranteed that the away side gets an early goal against.