Congratulations to IG for taking the top spot in the Dota 2 Asia Championships of 2017!
For a while now I've been thinking about doing an analysis of the top teams warding strategies to see if there are any potentially exploitable patterns. With DAC over, the Kiev Major coming up soon and finally some spare time, the runes have aligned.
The value of wards and dewarding
Vision plays a huge role in Dota, but it's difficult to quantify. You can't really put a value on the vision provided by a ward in a particular location at a particular time in the game. Dewarding is just as complex, except at least we know that the net gold cost of a successful deward is 0g, and a failed one is 100g. What interests me the most is this: if we can calculate the likelihood that teams will place wards at known locations and times, at what percentage does the benefit of eliminating the ward outweigh the potential 100 gold loss? 90%? 50%? I'm happy to leave that up to the pros to decide, but I personally believe dewarding is significantly underrated and should have a much larger focus, especially at the professional level.
I've broken the analysis into three parts: starting wards, early game, and later. Each period is then broken down into wards placed when playing Dire, Radiant, and a combined view. I've added my own amateur analysis and some stats for each section, as well as some potential suggestions for taking advantage of the patterns that appear. I don't claim to be a great player, so please liberally apply salt and independent thought when reviewing these. Also, the sample size is fairly small, particularly for Dire, so keep that in mind as well (21 games total, 8 Dire - 13 Radiant).
Playing on Dire
It's immediately obvious that IG strongly favours vision for OP in the middle lane for the first set of wards. Only one game played on Dire didn't have a ward overlooking the area surrounding the river at middle. Most notably, in 62.5% of games, the placement was in the same location for dewarding purposes. The other ward is split fairly evenly between top and bottom radiant jungles, and I wasn't able to see any obvious correlations between their heroes/lanes and these placements.
Playing on Radiant
Once again we see IG's strong preference for vision at middle at the start of the game. This time wards are more diverse in location, with only 2 overlaps in 13 games. But the other ward tells a different story. In 100% of games on Radiant, IG warded top, middle or both. Not a single ward was placed in their own jungle or the Dire offlane jungle for the first 6 minutes. Also, in 3 out of 13 games on Radiant, IG opted to only place a single ward, and hold on to the second one until the minute 4 after the horn.
In 90% of games IG had a ward around middle at the start of the game, and 80% of the time it was on the opposing side of the river. The other ward was usually covering one of the opposing side lanes; if IG played Dire it was a fairly even split but if they're on Radiant it was always placed top. In only 3 out of 21 games IG had vision of one of the active runes. And despite the fact that in every game except one their opponents were carrying sentries at the horn, IG was only dewarded once, by NP on one of the two wards overlooking the bottom active rune.
It's worth noting that Boboka is the man in charge of the early game wards. 90% of wards placed before the horn were his doing. During DAC he mostly played Monkey King and Earth Spirit, both of which have move speed abilities at level 1 that he usually skills (Rolling Boulder 100%, Tree Dance 75%) and takes advantage of immediately to get wards out without being scouted. When playing on Radiant, his most common strategy is to smoke, TP or walk to mid river, ward the opposing high-ground and then go top to regroup with the rest of the smoked team for engagement at Dire's top bounty rune. On Dire, he will either follow the same pattern, or TP to the Dire bottom T1, ward bottom, and rotate through middle to ward the Radiant high-ground before heading top.
- If IG is playing Dire, drop a sentry on the Radiant side of the river just south of the middle lane. 62% of games is far to often to not take advantage of.
- If IG is playing Radiant, consider an early gank on their safe lane. If it's viable before minute 4, TP to the bottom Dire shrine and engage via the Radiant forest. Smoke optional, it's almost certainly not warded.
- Consider having a player TP mid immediately at the start of the game to try and scout Boboka. He is usually following one of three patterns and if you can observe which you can be fairly confident of the general whereabouts of IG's starting wards.
- If IG is playing Radiant, be very careful around the top Dire bounty rune at the horn. They will attempt to ward it before you arrive, and in almost every game they also had 3 players smoked on top of it to gank, which led to first blood on multiple occasions.
Early game (4min - 11min)
Playing on Dire
IG is still fond of vision in the middle lane in the early game. However instead of warding the opposing side as they prefer at the beginning, they tend to ward their own high ground. Of the 50% of games where this occurred, all wards were placed in the same spot for dewarding purposes (overlooking the top rune). The rest of their warding in this period is aggressive, almost entirely deep within the enemy jungle, and is very well spread with only two overlaps in 17 placements.
Playing on Radiant
A similar story to Dire. Strong preference for mid vision, with more on their own side than the enemies. With more data for Radiant we see some patterns occur with the remaining wards. The main entrance to the Radiant jungle is very popular, appearing in 46% of their games. The other most common areas are around Dire's bottom T1, the back of the large camp in the offlane jungle, and the ramp below the top T2.
There's two significant differences between IG's starting and early game wards that stand out to me. The first is focus on active rune vision, from 14% at the horn to 71%. This makes sense, OP has his bottle now and can more efficiently utilize the runes. The second is the depth of the wards. I don't have another team's analysis to compare to, but in my eyes they seem particularly deep for so early in the game. Dewarding was only slightly better at this stage of the game, 9 of 61 wards placed by IG were destroyed before expiring.
- Try to scout and eliminate IG's ward covering active rune in the early game, particularly if they're playing Radiant. Since they never ward their jungle or the bottom rune at the start of the game, I would suggest getting your third observer onto the plateau at the Radiant jungle entrance before the 4 minute mark, which would scout all three of their preferred ward locations in that area. If they don't ward there, target their highground between the top rune and mid lane.
- Smoke as close to the base as possible. IG's deep early game wards are likely to spot a lot of smokes that are initiated in the jungles.
Mid to Late game (11min+)
At this point there's enough variables in the game that detailed analysis becomes difficult. However I think there's still a lot of value to be gained by reviewing the following images. Professional players usually have good instincts when it comes to being spotted by wards, but with additional info on where their opponents have chosen to ward in the past I think they'd have a better chance at successfully dewarding.
Playing on Dire
Playing on Radiant
I've circled the hotspots I think are most important. Most of these won't come as a surprise; I'd expect to see a high density of placements around Roshan and on the plateaus from any team. The one standout for me is the stairs up to the Dire offlane jungle: when IG plays Radiant it was warded 9 times in 13 games. Also of note is the position of wards at Roshan depending on which side IG is playing: on Radiant they usually choose the edge of the river, but on Dire the plateau is the most common. Team efforts to deward IG in the mid-late game were quite a bit better, 35% of wards were destroyed before expiring (up from 14% in the early game).
One thing I couldn't help but notice through this review is that OG does a significantly better job of dewarding than any of the other teams. Across the full duration of the games, OG's average deward rate against IG was 38%, but the average for the rest of the pool was 23%. I strongly suspect this has something to do with 7ckngMad 😛
I hope you found this interesting, I really enjoyed putting it together and I'm planning to do a few more soon. If you've got some comments or feedback join the discussion on Reddit. And if you happen to be from a team or organisation that's looking for someone to do Dota stats/analysis work, please shoot me an email, I am very interested in working with you.
PS: If you're part on an amateur Dota team and you're looking for scrims, register here on challenger.gg with the times you're available and start scheduling matches now!