Archive for the ‘Human-Computer Decision Making’ Category

It’s been about a week since the conclusion of the World Cup.  Congrats to Spain.  Espana es campeones del mundo as the result of a beautifully played tournament.  My family has close ties to Spain, so as the towers of tortilla, piles of paella and streams of sangria have finally subsided, I’m happy to have shared in the joy.  I’m also left with a strange emptiness as the incessant buzzing of the vuvuzela subsides and my tinnitus becomes audible again. But, I digress…

I want to talk about a few interesting aspects of decision making that struck me throughout the tournament, not all directly related to futbol.

First, the technology controversy.  Take a look at the video  montage to illustrate what had so many frazzled about the real time decision making of the refs.  There is probably no more pronounced case than with the first clip of the disallowed goal in the England v. Germany game, though many believe Karma may have been the culprit.

Here is what Fifa president Sepp Blatter said in the midst of the controversy

No matter which technology is applied, at the end of the day a decision will have to be taken by a human being. This being the case, why remove the responsibility from the referee to give it to someone else?

It’s an interesting perspective.  At first glance it seems to suggest that whomever the decision maker, at whatever vantage point, in whichever time frame is subject to the same uncertainties and asymmetries of information to assess the decision.  This is clearly not the case, just look at the video.  Then, I started thinking maybe Mr. Blatter had a point? Reviewers are indeed subject to their own set of biases and perception issues.  In some cases “going to the tape” can fail to resolve our doubts.  In the most ambiguous cases this seems to have only furthered speculation and spurred new interpretation of the events no matter how many efforts are made to unravel the confusion and quiet the controversy.

It’s important to recognize the different kinds of decisions, the snap decision, and those where we can take the luxury of process and deliberation.  Both are necessary, and there are arguments for both the snap and deliberation.  When time pressure constraints require us to make snap decisions, or when faced with a poorly described objective or poor information in deliberation, we are always prone to being misled by our judgment.  Many of you may know the selective attention test.  If you do I encourage you to watch it again and try again to achieve the objective.  If you haven’t seen this before watch and enjoy.  You have about a 50/50 chance of being very surprised.

We need to apply judgment and carefully interpret our perceptions to make good decisions.  So is it snap, or deliberate? I say YES.  I find myself uncomfortable with the Sucker’s Choice that these debates often set up.  It seems we’re often unwilling to wrestle with the more complex questions of how to combine such approaches, rather than choose between them.  There seems to clearly be a role for “thin slicing” information to be used in the process of deliberation to avoid analysis paralysis.

So, there is drama in decision making.  When it comes to futbol, many argue that this may fuel interest in the game.  It makes it more than a sport about the pure superiority of one team over another and enters in elements of chance, destiny, triumph over adversity, and the risks of being toppled by fate.  We seem to fight against this.  Maybe it hits too close to home.  So when fans of a sport say leave it alone, maybe they recognize something.  Maybe they recognize that our decisions are riddled with uncertainty and aren’t perfect, and very likely and simply cannot be.  We can only work to reach that threshold of certainty that lets us act, and we have to find the means to efficiently and effectively decide given the circumstances, and then have the courage and character to deal with the consequences.

The more we can come together and rely on each other to try to make sense of complex situations, recognizing technology as an extremely valuable ally (with imperfections) that we can partner with to better focus the picture or aggregate our judgment, we can often times improve our outcomes.

Luckily, if all else fails there is a fall back.  We can always submerge our options in an aquarium and drop in an Octopus to sort things out.

Read Full Post »

Being a strategist, I have a fascination with chess, or at least the ideas of chess. It conjures images of fierce competition, intellectual rigor, intense strategic thinking and steely eyed focus.  It can also be surprisingly dramatic and controversial, with colorful characters.  I really want to be able to play well.  I can beat my six year old son fairly handily, his brain is not yet fully developed, so this should give you some insight into my skill level. I have several iPhone apps that are fun and instructive and provide a useful distraction and brain exercise to combat the monotony of tarmacs and airports.  I like to see me and my iPhone as my personal rendition of the legendary competition between undisputed World Chess Champion Garry Kasparov and IBM’s Deep Blue Chess computer, except in my melodrama the opponents both have less capacity and the skill gap is greater and skewed to the machine. My level of pure frustration with chess is very much on par with what my friends who play golf describe as their emotional relationship to that sport.  One good move or a flash of insight keeps me coming back to the chessboard like a good approach shot does them to the fairway.

Garry Kasparov is considered by many (especially in the post Fischer era) to be the greatest chess player in the world.  He wrote an interesting book called ”How Life Imitates Chess”, which in the end is very much a book about… decision making.

Some of my own progress as a chess novice has been stunted by analysis paralysis.  Determining options can be daunting, choosing which to pursue can be even more so.   After only three moves the number of possible positions on the board can be well over 60,000.  So how do we decide?  If this were a purely analytical process based on logic and analysis, it seems that when Garry Kasparov faced IBM’s Deep Blue computer in 1996 and 1997 that the ability of the computer to win these matches should have been a foregone conclusion as it is when I compete against my iPhone.  Yet Kasparov won the 1996 match 4-2, he lost the 1997 rematch narrowly 2-1/2 – 3-1/2.  He offered to play a third match during an appearance on Larry King Live with a number of conditions, including a willingness to concede Deep Blue as world champion if it won the match.  IBM chose not to take him up on the offer despite the computer’s ability to calculate 200,000,000 positions on the board per second!

There are four basic chess values that Deep Blue must consider before deciding on a move. They are material, position, King safety and tempo.

Material is easy. The rule of thumb is that if a pawn is considered to be worth a value of 1, pieces (knights and bishops) are worth 3 each, a rook is worth 5, and the Queen 9. The King, of course, is beyond value, since his loss means the end of the game. This varies slightly in certain situations — retaining the Bishop pair in the end game generally increases their value beyond 6, for example – but the laws of material are fairly constant.

Position is more complex. In the old days, it was thought that control of the center was all that mattered. Nearly all grandmaster games before the 20th century began with Pawn to King 4 or Pawn to Queen 4. Control of the center is still important, but certain grandmasters in this century found some effective “hypermodern” openings that delay development of the center, with the idea that the opponent will overextend his position and leave himself vulnerable for attack.

The simplest way to understand position is by looking at your pieces and counting the number of safe squares that they can attack. The more squares they control, the stronger the position. Thus, a seemingly quiet pawn move can be very strong if it opens many new squares for a more powerful piece behind it.

The defensive aspect of position is the safety of the King. This is self-explanatory. A computer must assign a value to the safety of the King’s position in order to know how to make a purely defensive move.

Tempo is related to position but focuses on the race to develop control of the board. A player is said to “lose a tempo” if he dillydallies while the opponent is making more productive advances.

The programmers have defined how Deep Blue’s program evaluates these factors. The computer then searches through all the legal moves and chooses the one that yields the highest value.

I don’t know about you, but this process is beyond my computational capacity in any meaningful and constrained time frame.

So let’s think about this.  It’s been estimated that through a process of elimination and prioritization of high potential moves, that human chess masters consider approximately three dozen serious options or so before making a move versus two hundred million per second by Deep Blue.  Then, we apply a mix of analysis, judgment, preference, creativity, experience, intuition and a dash of guts to form a decision cocktail and make a choice (or sometimes take a gamble), often resulting in a very good outcomes.

So, How do Garry Kasparovs work?

So is this yet another tale of man versus machine, like the folklore of John Henry versus the steam hammer? Is it better to have nearly limitless computational capacity with limited experience and intuition, or vast experience and intuition and less computational capacity? It seems to me that a Deep Blue Kasparov would be invincible. Maybe man versus machine is the wrong question?

Read Full Post »

%d bloggers like this: