- In a wide range of problem-solving settings, the presence of a familiar solution can block the discovery of better solutions (i.e., the Einstellung effect).
- To investigate this effect, we monitored the eye movements of expert and novice chess players while they solved chess problems that contained a familiar move (i.e., the Einstellung move), as well as an optimal move that was located in a different region of the board.
- When the Einstellung move was an advantageous (but suboptimal) move, both the expert and novice chess players who chose the Einstellung move continued to look at this move throughout the trial, whereas the subset of expert players who chose the optimal move were able to gradually disengage their attention from the Einstellung move.
- However, when the Einstellung move was a blunder, all of the experts and the majority of the novices were able to avoid selecting the Einstellung move, and both the experts and novices gradually disengaged their attention from the Einstellung move.
- These findings shed light on the boundary conditions of the Einstellung effect, and provide convergent evidence for Bilalic, McLeod, & Gobet (2008)’s conclusion that the Einstellung effect operates by biasing attention towards problem features that are associated with the familiar solution rather than the optimal solution.
- Paper suggested by Dave.
- See – free access – Link
- See also Scientific American Link – unfortunately, the full article isn’t open access, so without a subscription you can only see a summary.
- For a brief Wikipedia article on the Einstellung effect, see Link.
- The Einstellung effect is also mentioned in "Smith (Alastair D.) - Thinking and Problem Solving".
- For a 2-part discussion on ChessBase1 ("Sheridan (Heather) & Lahaye (Rick) - Psychological warfare and Einstellung effect") see:-
- The first part of the above article includes a helpful link to "Bilalic (Merim), McLeod (Peter) & Gobet (Fernand) - Why Good Thoughts Block Better Ones: The Mechanism of the Pernicious Einstellung (set) Effect" (Link), one of the earlier studies referenced by the original paper – and that referenced in the Authors’ abstract above – on which it claims to make an advance.
- In turn, the above paper frequently references "Bilalic (Merim), McLeod (Peter) & Gobet (Fernand) - Inflexibility of Experts – Reality or Myth? Quantifying the Einstellung Effect in Chess Masters", a longer version of the same.
- There is a more recent – and much shorter – paper by the same authors ("Bilalic (Merim), McLeod (Peter) & Gobet (Fernand) - The Mechanism of the Einstellung (Set) Effect: A Pervasive Source of Cognitive Bias"), but it seems to be milking the same general theme.
- I don’t think I’d heard of the Einstellung Effect prior to this investigation. I thought I had, but it was just the word Einstellung (German for “attitude”, in this context) in "Winch (Peter) - Eine Einstellung Zur Seele".
- There is much folk psychology in all this which suggests contradictory nostrums. There are two sayings:-
- The good is the enemy of the best, which relates to the Einstellung Effect (see for example Link), but also
- The best is the enemy of the good – see Link - which relates to the law of diminishing returns, “polishing the rocket” and all that.
- If something’s worth doing, it’s worth doing well, and
- If something’s worth doing, it’s worth doing badly - ie. if it’s urgent and there’s no time for perfection.
- Where the Einstellung Effect advances on these folk nostrums is in its identification of a cognitive blockage, as distinct from simply misdirected, or inadequately directed, effort.
- What of the relative merits of all these papers?
- My view is that the papers by Bilalic et al are the better-written and technically less open to objection. Their examples are of traditional chess problems, and the presence of a familiar but sub-optimal solution masking an optimal, less familiar solution to the extent of making an expert chess-player perform at a level 3 standard-deviations weaker is striking. Even with the very strongest players there is some impact on the time taken to find the optimal solution.
- The papers by Sheridan et al, while claiming to make advances (a) technically – in focussing on “dwells” and (b) in the problems themselves – by showing that the Einstellung Effect disappears if the Einstellung move is a blunder seem open to a technical objection in the set-up of the puzzles.
- While the problems do not ask for “mate in x”, they are set up in such a way that the solver is induced to think that there is such a solution, which may explain why the solver’s focus remains on the indicated area of the board. The cover is blown somewhat when the Einstellung move is a blunder, and the solver may quickly become resigned to the thought that this is just a “best move” problem, as over the board, rather than a “neat and tidy” checkmate problem.
- That said, in real life – at least in over-the-board chess playing – it is usually unclear whether there is a quick and neat solution. Remember the maths exam questions – where if you got a scruffy answer with a string of decimals, this was an indication that you’d gone wrong, while in real life there’s no such indication.
- So, it may be that the Einstellung Effect operates by a fixation on the thought that the problem before us has this “quick and neat” property when in fact it does not. I have certainly myself wasted lots of effort over the chessboard looking for such quick wins that are in fact not there.
- The four problems are to be solved – or at least a decision is to be reached – in 3 minutes each. My choices fell into the categories of “optimal moves” according to "Sheridan (Heather) & Lahaye (Rick) - Psychological warfare and Einstellung effect" (Part 1), but were uniformly the second-best options according to "Sheridan (Heather) & Reingold (Eyal M.) - The Mechanisms and Boundary Conditions of the Einstellung Effect in Chess: Evidence from Eye Movements". Ie:-
I think the reason was that I was still assuming that, Einstellung moves aside, this was still a “mate” problem, so was making moves to increase the pressure in the critical area, rather than simply making the objectively strongest moves as a computer would.
- Nc2 (rather than Ng2)
- Rf3 (rather than Na3),
- b32 (rather than Rg5), and
- Rd1 rather than Rb3.
- Irrelevant Comment: The same authors have something interesting to say on sample sizes and extreme outliers in "Bilalic (Merim), Smallbone (Kieran), McLeod (Peter) & Gobet (Fernand) - Why are (the best) women so good at chess? Participation rates and gender differences in intellectual domains".
Footnote 1: Heather Sheridan is one of the authors of the original paper, on which Rick Lahaye was a Chess consultant. It seems that Rick Lahaye is a FIDE-master-level chess-player.
Footnote 2: Well, actually, I chose the more aggressive c4, which is at least preferred to b3 by the computer!
Text Colour Conventions (see disclaimer)
- Blue: Text by me; © Theo Todman, 2018
- Mauve: Text by correspondent(s) or other author(s); © the author(s)