# How likely is a comeback victory, how much does starting side really matter? A look into every map and region during Stage 2.

Learning to attack is the key on every map.

During the Berlin Major, the most one-sided map in the pool was Skyscraper, with a 61 per cent defender win rate. A 4-2 half would be a 66 per cent win rate, meaning that anything worse than 3-3 half is bad for the team behind, no matter how attack or defence-sided a map is.

These percentages could, however, be pretty misleading. A 50 per cent attack/defence split could be caused by a load of 6-0 attack wins in certain games that were cancelled out by 6-0 defence wins in other games. Alternatively, it could just be that a team that likes Skyscraper a lot is also very good on defense, and less due to the map itself.

Let’s dive deeper into the map win rates to see how likely teams are to come back from 2-4, 1-5, and 0-6 deficits on each side on each map.

Note: We’re only looking at the first half of each map as the second half can end early. For example, in a 7-3 result, the second half took just four rounds, meaning the two worst defensive sites weren’t played a second time, skewing the numbers. Additionally, if you only take in account completed second halves, you’re also giving an extra focus to close games and skewing figures again.

## Overall takeaways

In total, the team that starts on the attack win the game 54 per cent of the time. While that may not sound like a lot, that’s a bigger bias to one side than all but three maps during the Berlin Major.

When graphed, you can see clearly how likely the attack is to win on each map after each of the seven possible first-half results. You can find out the defence numbers by just taking the reciprocal of the attack figures. For instance, a 40% attack win rate after a 1-5 half on Theme Park equates to a 60% win rate for the defence after going 5-1 up.

As shown in the graph, a 3-3 half is overall an even game, with the attacking side winning in 49 per cent of games that hit the 3-3 scoreline. If the attack goes 4-2 up, it is very unlikely (13% chance) that they will lose the map, but this is not always the case on the defence (34% chance). On Theme Park, Kafe, and Oregon, the defence won just half of the rounds in which it led by 4-2 at the half.

It's worth clarifying that the higher the numbers in the table above, the harder it is to attack. This is as Skyscraper, for instance, saw 100% of 3-3 scorelines be won by whoever was attacking. This shows attacking is hard and simply winning half your rounds is a massive victory. On the other hand, even when you win five attacks on Border, it's no guarantee you can get just two defences. Hence low percentages mean that defending is hard.

Looking at the exact figures, it’s remarkable that no team that started on the defence across Stage 2’s top-flight leagues, or the Berlin Major, won Skyscraper without taking a 4-2 lead. The side that started on the defence also needed at least a tied half on Villa, Theme Park, and Kafe to win the map; they could not come back from behind.

The same can be said for Border on the attack, as a minimum of a 3-3 half was needed to get a win here.

Caption: See this in full definition here.

Across all seven tournaments, the defence never came back from a 0-6 deficit and only once did it come back from 1-5 -- this was when Soniqs mounted a comeback against SSG on Border on play day five.

On the other hand, the attack had better luck. On seven occasions, a 1-5 offensive half was salvaged thanks to a better defensive half.

Overall if a team loses the first half, they have an 18.46 per cent chance of mounting a comeback; 11 per cent if that deficit came on the defence and 25 per cent if on the attack.

In total, all these statistics tell the same story, if you do badly on the attack, it can be salvaged. If you do below average on the defence, it’s probably game over.

It also allows you to rank the maps by how “defender-sided” it is by looking at how likely it is for a team to win their half rather than judging it site-by-site. This gives a wider look at the map as a whole and helps stop a single easy-to-defend objective skewing overall numbers. For example, if Villa’s Aviator has a 100 per cent win rate, but Statuary and Kitchen are at 20 per cent, it may look even overall. However, in reality, it’s not a great map to defend.

Let’s focus on some of the peculiarities, tournament-by-tournament:

### Only once in the whole of the Berlin Major did the defence lose the first half but still win the map

On Border, between DWG KIA and Rogue, the eventual champions went down 2-4 but clawed the game back to overtime to win the map.

That was the only case in the entire major when the team that began on defence didn’t need to win three of their first six rounds to win the map.

76 per cent of 3-3 halves were won by the team attacking first in Berlin

25 of the 68 maps played saw a 3-3 scoreline. Out of these, the team which attacked first won 19 times, compared to just six times for the defenders.

This is because it is generally easier to defend, so a 3-3 half is all that’s needed to give a massive advantage after the swap.

Caption: See this in full definition here.

### In Berlin, Bank, Chalet, Kafe, or Villa were never won by the team that lost the opening half

The same is true for Bank, Border, Oregon, and Theme Park in the EUL, Chalet, Kafe, and Oregon in the NAL, and Bank, Border, Clubhouse, and Kafe in the BR6.

Caption: See this in full definition here.

### Chalet was a messy map in Brazil

Chalet was played 10 times in the BR6 League and won by the team starting on the attack six times. Out of these six, the attack had finished their half with the following scorelines; 6-0, 5-1, 4-2, 3-3, 2-4, and 1-5.

This led to the odd occasion where the attack won 100 per cent of maps where they started off 1-5 down, but only 50 per cent where they finished 4-2 or 3-3.

Caption: See this in full definition here.

It then became much cleaner at the Copa Elite Six:

Caption: See this in full definition here.

### In the 15 maps where the defenders lost the first half, NA came back three times

For comparison, the EUL saw just one comeback in 13, the BR6 had 1 in 15, and APAC had two in 19.

Caption: See this in full definition here.

## APAC South had the largest win rate for whichever team starts on attack (61 per cent)

While APAC North was in line with the other regional leagues, APAC South strongly favoured the coin toss.

The biggest difference between these two leagues was likely who won after a 3-3 tie. In the South it was the team who attacked first 62 per cent of the time, while in the North this fell way down to 14 per cent.

Caption: See this in full definition here.

You can view all the above information on Google Sheets here.