Welcome to my eighth and final Woodpecker Method training blog. It was the 25th of July when I started on this journey, with the initial aim of solving 984 exercises (comprising all 222 Easy and 762 Medium exercises in the book) over a 28-day period, followed by repeated cycles with the final goal of solving all 984 within a single day.
As usual, I’ll start by recapping my results from previous cycles.
1st cycle: 1033 minutes; 88.9% accuracy
2nd cycle: 663 minutes; 93.7% accuracy
3rd cycle: 366 minutes; 98.7% accuracy
As I explained in last week’s blog post, I decided to adapt the schedule in order to finish early and give myself time to recover before the Batumi Olympiad. So, rather than three more cycles lasting 4 days, 2 days and 1 day, I went for two more cycles, lasting 3 days and 1 day respectively.
Cycle 4
I solved all exercises in a combined 267 minutes. I was happy with this time, which knocked 27% off my time from the previous cycle. I didn’t write down my answers or keep score this time, for a couple of reasons:
* In my third cycle I was already close to 99% accuracy, so there didn’t seem much point in tracking what could only have been a small improvement.
* I already knew most of the solutions quite thoroughly, so it seemed logical to skip writing/checking answers in order to save time.
There were still a few positions where certain details of the solutions remained unclear in my head. Whenever that happened, I noted the number of the exercise and my answer, then checked the solution at the end of the session.
Cycle 5
The big one! Solving 984 exercises in one day sounded daunting at the beginning, especially after reading Axel’s story on page 9 of the book about spending 22 hours in a basement! So, it feels a bit anticlimactic to say I was able to work through the 984 exercises in a combined 198 minutes – almost a 26% reduction from my time in the fourth cycle. Make no mistake – it was still a grind, but breaking the task down into smaller chunks helped.
My score for the 222 Easy exercises was up to 100% by my third cycle, so by the fifth I knew the solutions well enough to get through that section in something between 20-25 minutes. For the Medium section, I calculated that seven mini-sets of 108 exercises would see me through (with an extra 6 positions at the end). So I solved 108 at a time, took a break and then repeated until the job was done.
A while after completing the set of 984, I couldn’t resist tackling another 16 from the Advanced section, just so I could say I did a thousand in a day. I scored 75% but didn’t keep track of the time. More importantly, I found a duplicate exercise from the Intermediate Section, so we already have a replacement ready for when we reprint; and we will of course inform our partners at Forward Chess and Chessable.
Final Thoughts
I’m happy I worked through the Woodpecker Method: I feel much sharper tactically, and will feel confident in Batumi knowing I have this amount of training behind me.
Although the training was tough, I can’t help feeling that I could/should have made it even more challenging. Axel spent 22 hours on his last cycle whereas mine was much quicker – and I was still checking variations and calculating, rather than just seeing the first move. I can’t imagine my tactical ability is much (if at all) better than Axel’s, so I can only assume he was working with a more challenging set of exercises – so maybe I will look at doing the same at some point in the future.
For now though, I will focus on the Olympiad; hopefully my improved tactical sharpness will show in my results. For future chess training, I’ll take the emphasis off tactics for a while. I’ve been working on some endgames over the past few days, and I may well try ‘Woodpeckering’ Jacob’s books Positional Play and Strategic Play at some point, though probably not for as many cycles.
Thanks for reading this blog series and, as usual, please do share your training results and questions in the comments. Although this is my final Woodpecker blog, we will have a final Woodpecker Wednesday next week with Axel himself as the guest contributor!
I finished my 3rd cylce doing the 984 exercises set today on schedule.
Focussing on accuracy, but not forgetting to improve speed I managed to get to 95% accuracy (i.e. making only half the mistakes compared to cylce 2) (not great, but workable), while reducing the total solving time from 1280 min to 684.
The missing accuaracy is mainly due to missing some details, the 1st moves are almost always correct, but sometimes I went to much for speed and messed up.
Still happy with my progress and motivated to continue, but some work left to be done. I will do a little longer break due to traveling next weekend and tackle the next cylce next week, probably from Mo-Thu.
And it is encouraging to hear that the full day is a manageable exercise without a locked basement ;), even if I will not take the shortcut like Andrew.
@tarsitius
Well done for making such a big jump in speed in your third cycle.
You mention that I took a shortcut, but remember my results for that final day would have been even better if I’d followed the program exactly as the book recommends! With speed and accuracy improving with each cycle, it’s harder to do the single-day solving on the 5th cycle rather than the 6th. As I said before, I only did it this way to give myself a decent break before the Olympiad.
@Andrew Greet
Thx.
I did not intend to critize your shortcut-
but I am impressed with your achievment & wish you a succesful Olympiad!
@tarsitius
Thanks for the good wishes! Yes I didn’t take it as criticism – it’s just that the word ‘shortcut’ has connotations of something we do to make things easier for ourselves. So it was just to remind anyone else reading that reducing the number of cycles has the opposite effect; I was able to get away with it and still finish comfortably, but I want to make it clear to everyone that it is preferable to stick to the program.
This training method looks effective…as shown by your results : better ratio of solved exercices and an overall improved speed.
I understand that by doing the same exercice again and again, you will train to solve this type of position on a more automatic way. But what is the true progress ? you do it better & faster because you already knew the solution , not because you are better at tactics, ?!
I think you should keep about 100/200 diagrams from the original set apart from the training phase. Then at the end you solve 100/200 diagrams from the training program and reccord your performance ( ratio and time) .then you do the same with the 100/200 “unknown” diagrams . If you get the same score ( or close) with the 2 sets of exercices then it is ok.
what do you think about ?
After my last post on my 3rd cycle last saturday I decided to follow in Andrew´s footsteps by skipping the free day.
So I started my 4th cycle on sunday when I was able to solve 318 exercises followed by 96 exercises each on monday and tuesday. I still wrote down the first move of my solution for the intermediate exercises but only compared it when in doubt. That didn´t happen too often since by now I know almost all soluitions by heart.
Therefore I also didn´t keep track of my score anymore. The three sessions summed up to almost 6 hours which was much less than I expected.
I will try my best to do the 5th cycle on thursday and friday and sunday is reserved for the final one.
@RYV
This is a question which will be answered when I get to play some OTB games after the woodpecker training.
Of course some part of it is remembering solutions, but as I try to solve the exercises every cylce by working out details I do not remember- as recommended- and not just to remember- I am pretty confident that I will see some improvement in tactical ability.
After all the full programm will require – very rough estimation- 60hrs of serious training- and I am confident already that it is more effective for me than spending 60hrs training with a tactics trainer tool or even more effective than studying some openings.
no doubts that hard training will improve your abilities.
My question is more about claiming the method works because at the end we are able to solve more diagrams & we do it faster. Why not just spend the same amount of time on new puzzle each time ? you should also be able to improve on ratio and speed without the “already knowing” question.
If you recognize some patterns, you drive through the city with a map.
If you recognize a huge amount of patterns, you drive through the city with a navigation system.
“Why not just spend the same amount of time on new puzzle each time ?” If you do that, I don`t think that you get the patterns into the long-term memory.
I think RYV’s suggestion is really interesting. In machine learning, we’d call that creating a hold out test set, and use it to validate that the learning that’s happening generalizes well to previously unseen problems.
Hey
I have completed my first cycle and looking forward to complete the full 7.
I was wondering if some people have already demonstrated an improvement ? That should be pretty easy by checking the online blitz rating evolution, has anyone see progress in this area ?
Tnx
@RYV
It’s perfectly reasonable to want to measure the effectiveness of the training in some sort of scientific way. How exactly to do this is more complicated. I’m confident that the training has improved my tactical ability, but testing this rigorously would require a proper study involving large numbers, test/control groups, statistical analysis and so on.
Let me answer your question about claiming that the method works because of the improved speed and accuracy. I don’t see my training result as a successful outcome in itself. My training was successful in the sense that I was able to achieve or surpass the targets I set for myself – but I only count it as a true success if it makes me a stronger player over the board.
As for why this method (hopefully) works better than solving puzzles without repetition, the theory of it is that all 984 of the positions I’ve solved contain a tactical motif relating to the way certain pieces are configured in a pattern. If any one of those patterns occurs in one of my games, I should recognise it at a subconscious level, because I have drilled the motif into my mind through this repeated solving.
I will look at my results at the Olympiad and subsequent events as an indicator of how well the training has worked.
Cycle 1 results: 40 hours, 87%
Cycle 2 results: 18 hours, 92%
Cycle 3 results (so far): 335 problems, 2.75 hours, 94%
Just started cycle 3 on chessable (!) yesterday as I was invited to Beta test. I’m not going to lie, my first session was very rocky – I didn’t realize that seeing all of the Black positions from Black’s perspective would throw me for such a loop. A few I actually looked at in the book instead. I tried to make up the extra time on the white problems and missed quite a few.
I calmed myself down and did another session, not worrying about time as much, and it went much better in terms of accuracy and it went relatively quickly too.
So far it seems reasonable to do this cycle in 7 days and be set up for a doable 4 day cycle. We’ll see though, I certainly am not experiencing a magical 99% accuracy and automatic solving on all the problems yet. 😉
Well, that is a very good question and is discussed in the intro to the book. Ultimately neither author claims it for sure would be. However, I think a big reason Axel was drawn to this method is because he knew two players who were making a big push in studying tactics – Tikkanen and another player. The other one didn’t achieve big results until after he started doing the *same* tactics over and over – rather than constantly looking at new ones.
Given that each cycle goes faster, you only would be able to do ~2x as many tactics if you just did more tactics. It’s not hard for me to believe that doing ~1000 tactics 6-7 times is going to lead to a bigger difference in my tactical ability than doing ~2000 tactics 1 time.
Despite Andrew’s advice I did the third cycle of 473 exercises in one go today. My progress is as follows:
Cycle 1: 473 exercises in 995 minutes, score: 91,5%
Cycle 2: 473 exercises in 580 minutes, score: 95,1%
Cycle 3: 473 exercises in 275 minutes, score: 98,1%
So again I increased my speed with a factor two, and accuracy went up substantially as well. Since I already did the whole cycle in one day and both my speed and accuracy are already high, I’ll stop here. Maybe in the near future I will start with the remaining exercises but for now I’ll focus on other aspects of the game. Of course the real proof of the pudding is in the eating, so I’m I’m expecting a spectacular increase in rating or I want my money back 🙂 .
@Ray
Start playing the Modern Benoni and the Najdorf Poisoned Pawn for the best results. 😉
I am surprised that allmost everyone Start with a Score around 90% . It just gives you à potential 10% Progress . Not much!
@RYV;
However, this consideration does not take into account the time factor. For example, I needed several minutes for many solutions. I had to work hard for them. In a game over the board I do not know if there is a solution. If so, I might need to invest a lot of time to find the solution. In such a situation it makes a big difference if the solution pops out of the subconscious mind within seconds, especially in time trouble.
I’ve continued woodpecking my way through the easy problems. I’m four rounds in now and am being quite a bit more relaxed about the schedule than the official recommendation.
That said, I do seem to be playing much better tactically. I just beat a GM in an otb simul, which I’m completely sure wouldn’t have happened without this improvement.
Andrew, good luck in the Olympiad!
I still believe the gain in time is due to memorisation from one week to the next one, getting the pattern at subconsious level will require more than a few repetition. But i agree that the goal of this method is to transfer your ability in tactics from the calculation step to a more automatic level. Anyway, in time trouble situation, this will Not help you much as it is more important to avoid blunders ,which is to me completely different than solving a diagram situation.
@Ray
Well done for completing your set in a day. I hope you’ll be able to play in some tournaments soon and experience an improvement. Regarding any money-back guarantee, surely you’re aware of our policy that if you play great after working with the book then the publisher takes the credit, and if you don’t then it’s all your own fault! 🙂
Joking aside, I hope you found it useful. Perhaps after training some other parts of your game you’ll come back and attempt a more ambitious Woodpecker set, say in six months or a year’s time. If you were able to finish this set in a single day on only your third cycle, it shows that you could have managed many more exercises if you went all the way up to 6/7 cycles.
@RYV
Hopefully 5/6/7 cycles is enough to drill the patterns into the subconsciousness.
Regarding performance in time trouble and the related topic of blunder avoidance, Hans mentions this in his introduction as one of the main areas in which Woodpecker training benefited him.
@Andrew Greet
Thanks for your comments! I agree it shouldn’t be too difficult to do the full 7 cycles with all the exercises – the reason I didn’t do this straight away was primarily that I didn’t have that much time available in the first cycle. But that’s mainly a matter of priorities.
@Seth
🙂 The problem is that most white players are avoiding these two openings nowadays (or any other sharp openings, for that matter).
I´m finally done with woodpeckering too. 😉
My results of my last two cycles with a set of 510 exercises:
Cycle 5: 173 minutes
Cycle 6: 145 minutes
I no longer kept track of the points scored since I was at 99% before. Given my comments after the third cycle the speed went up beyond my expectations but probably RYV is right that memory played a big part here. However, even in the last cycle I forced myself to calculate and visualize all the impoortant variations.
I also did another set of 24 positions of intermediate level from the book that where new to me. Here I saw the right move quickly with one execption where I had to set up the position on a board. Working out the variations, writing them down and comparing with the book took up most of the time. I needed 65 minutes to score 98% overlooking only one important try in a sideline. Compared to the first cycle that is an increase of speed by around a third and accuracy also went up (usually 93% in cycle 1).
…
…
Of course this is not enough to draw any conclusions but it hints to me that I´m on a good way.
But the best thing I can say about the Woodpecker so far is that it brought me into a habit of working on chess on an almost daily basis which is something I will try to keep up for the future.
For this I will use the remaining exercises from the book (6 to 12 exercises a day) for normal tactial training but my main focus will be on revising and expanding my knowledge of basic endgames, something I haven´t done for years.
@Andrew Greet
Well, Not sure i will find the motivation to solve 300 pages of exercices 6 Time again..but at least i will read carefully the 25 first pages.
Well done to all of you who have made it getting the ultimate Woodpecker title. Andrew made me change my mind, so I will finish my 5th cycle today (2 days) with the 984 exercises. This cycle has had a dramatic impact on my speed and error ratio. I seem to recognize everything now which is fantastic for a relatively weak player! Tomorrow, I will try and go for the 6th cycle and perhaps also include 16 exercises to make it to the golden 1000. Will you also use the Woodpecker as some training before tournaments, where the exercises are solved relatively quickly?
22…Bh4
Wow.
Andrew, thanks for sharing your Woodpecker experiences with us. Today, I also completed it solving 1000 exercises this nice Wednesday. In the final cycle, my score was very close to 100% and the solving speed was very high. It will be extremely interesting to see if people in the Woodpecker Society get some remarkable results in the near future 🙂
Well done everyone who has been working hard on their training. My Olympiad results have been tough so far, with a win over a 2100 and losses to Giri, plus two other GMs in the 2600/2500 range. I feel sharper from the training, but guys like Giri are obviously in a different league. Today is a rest day. I feel my overall play has been decent, but the opponents on board 1 are so tough that there’s little margin for error. Hope to pick up the pace in the second half of the event.
@Andrew Greet
Good luck. I hope you will win a lot of games at the Olympiad. Yes, it must be a lot of fun and very tough to play board 1!