Marko Hurst
Analytics & GamblingHow Similar They Really
Are
Analytics made so simple my Grandmother knows what I do
now
Marko Hurst
Me
Book: Search Analytics - Conversations With Your Customers
• Anticipated release: December 2009
• Book website: RosenfeldMedia.com/books/SearchAnalytics
• Co-Author: Lou Rosenfeld
Consultant, Author, & Speaker
User Experience
Web Analytics
Neuroscience
Blog: MarkoHurst.com “Insightful Analytics”
Twitter: MarkoHurst
Contact: [email protected]
Marko Hurst
Purpose
To quote Eric Peterson “Web Analytics Is Hard”
• Leave the data analysis to the experts
• But what if you want to just be able to understand Analytics at a basic level?
• Then keep reading
Raise your hand if you’ve ever played a card game
• Blackjack? Poker? Uno? Doesn’t matter
• Then you have used, perhaps unknowingly, processes, rules, & data to optimize and improve your outcome (results)
In less than 10 minutes using card games / gambling as an analogy you should understand at a basic level how analytics works
Marko Hurst
Viva Las Vegas Baby!
Marko Hurst
All I ever needed to learn about analytics I learned in Las Vegas
Growing up in a Sicilian family meant three things…
1.Our cheeks would be sore from all the damn pinching
2.Great tan
3.Hearing Grandfather’s lectures on how to get ahead in the world by beating odds
The last point included more than lectures and eventually lead to learning more about gambling than a kid probably should
• I could count a deck down in less than 25 seconds (by age 14)
• I learned to count multiple decks, including side decks
• I learned multiple counting strategies (basic & advanced cards counting techniques)
Marko Hurst
So, before you think I grew up a
degenerate gambler (or a lush)… what
do cards have to do with Analytics?
Everything! Let me show you.
Marko Hurst
Before we get into details
Let’s start with the basics that never change
What’s the goal of gambling?
• To win, plain and simple
• Always keep your GOAL in mind
• It will tell you when to fold, raise, bluff, or run like hell
How do (should) you go about winning?
• Know the rules.
• Know how much you are willing to loose, a.k.a. Measure
• Watch the cards, dealer, other players, a.k.a. Monitor
• Apply strategies to improve your chances of winning (more), a.k.a. Optimize
All the above is true for analytics as well
Marko Hurst
Similarities of card games (gambling) & analytics
Card Games Attribute AnalyticsReal time Timing Real time, near real-time, just-in-
time, delayed
Cards (colors, symbols, values) Data Numbers, %’s, KPIs, metrics, etc
Multiple decks, high-level counting strategies, side decks, lights, noise, etc.
Data overload Too many data sources, no clear business or monetization model, etc
Running count, true count, knowing what cards have already been played
Prior Data Reporting
Improving your hand with more or less cards
Filtering Segmentation
Straight flush, full house, 3 of a kind, etc. Patterns Pattern recognition / analysis
Read, understand, comprehend, and decide what actions to take to improve your goal of winning
DecisionMaking
Data analysis
Card counting, side decks, etc. Predictive Modeling
Forecasting / prediction
Betting strategies, player strategies, etc. Strategy (Performance) optimization
Marko Hurst
Card games are to analytics…
1. There is a process and rules to govern it
Like any activity requiring a series of steps card games (Blackjack, Poker, etc.) are played in a specific order
• The game wouldn’t work if we first showed our entire hand, then placed our bets, and then dealt the cards
• Similarly, analytics is also a process guided by rules
Rules, or Process, is key for any card game
• The rules are clearly defined
• Dealer actions are pre-scripted
• Except for the few actions of placing a bet or collecting your winnings, everything is largely automatic
• Analytics is no different
Marko Hurst
Card games are to analytics…
2. We must make decisions with less-than-perfect information
• In Vegas, you have to gamble at every step of every hand… guessing how things are going to play out long before we’ve seen all the cards
• We frequently have to make important decisions about which data to use before we have all the information
• The same is true of analytics
Marko Hurst
Card games are to analytics…
3. A complete game is made up of a small number of elements
All the data is contained entirely within the 52 cards that make up the deck and shared symbols on them
• 9 number cards (2-10)
• Four face cards (Jack, Queen, King, Ace)
• Four suits (hearts, diamonds, aces, spades)
• 2 colors (red & black)
There is still an infinite variety of ways to play any game. Likewise in analytics, a small set of data points will represent an infinite number of problem solving options.
Marko Hurst
Thank You!Good luck in Las Vegas and with your analytics
Book: RosenfeldMedia.com/books/SearchAnalytics
Blog: MarkoHurst.com
Contact: [email protected]
Twitter: MarkoHurst