Using analysis to assist your Decision-Making

Ultimately we want our statistics and analysis to help us make more informed decisions. Ben Alamar a leading Sports and Data analyst wrote a book about how decision-makers should interact with stats. His short but detailed account is extremely useful and worth a read no matter what sport you are coaching or playing. In this post, we will outline how we can use our stats and analysis as a tool to assist decision making. If we trust our data to be accurate and detailed enough we might even let it guide our decisions. No matter what weight we give, certain principles must be followed when using the analysis to ensure that they genuinely support decisions and not misguide us.

Facts and Interpretation

It is important to deliver facts as an analyst, impartial information, that is a fair reflection of the game. However, simply producing numbers on a page or table isn’t much use if the receiver cannot read the information. All of the information that is delivered should be done with context to explain the value and relevance of the statistic. What is the point of telling a manager we’ve missed 10 shots? A manager can see that they need an explanation of what they are seeing. Is it that we are missing 7 of our 10 shots on the left-hand side and need to focus on the right-hand side. Maybe it’s a specific forward that is missing or that we are shooting under pressure. All of these are possible explanations for the outcomes which is what analysis is designed to do. By explaining the actions on the pitch we can better understand them and with a better understanding of the game's events, we can make better decisions when adjusting in an effort to sway the game in our favour.

Useful

Detailed and quality information. This rich data set can be used in a multitude of ways to help analyst’s coaches and players develop. We can identify incidents and trends. Build player and team profiles of ourselves and opposition teams and offer detailed feedback to players.

Insightful

Insightful information is an accurate account of a small data set. We would consider a small data set one or two games. There is not enough data to confidently identify trends however it may offer insights into players’ behaviours and skills that the coaches can address. This data is extremely useful as a learning tool but less so for creating team strategies and tactics.

Misleading

This form of information is the riskiest information and is the most likely to be misinterpreted. Inadequate data should easy to identify. However, with a lot of information, we may fall into a trap to think that a high volume of information will allow us to make inferences and identify trends. Although we may be able to identify trends with this information, if we do not ensure the quality of the information then there is a chance that the trends we identify are incorrect. If this is the case this will mislead analysts and coaches to make poor decisions.

Inadequate

This information is not useful to anyone and It is easy to see why. With little information and now way to ensure it is quality information. It should be easy to identify when our information is inadequate.

Adding Value identifying miss matches

When considering our opposition the advantage of analytics is to assist in identifying competitive advantages that either team has. These are what I describe in a simple terms “Mismatches”. Where do we have an advantage in the game that we can exploit?

As an analyst when comparing your team and an opposition team, a particular focus should be focussing on mismatches because these present unique opportunities outperform your opposition. Gaelic Games is a 15 vs 15 field sport. A lot of the positional, individual and tactical match-ups will be close. There might be two or three key battles where each team can win the game. As an analyst, it is your responsibility to identify these unique competitive advantages where you can win but also lose the game and assist in devising a strategy to deal with these to ensure they favour your team.

A mismatch either positionally or tactically can create favourable situations for your team. For example, you could be playing against a team that has a short full back line and thus a kicking game into an aerial superior full-forward line creates an advantage for a team and use this. If you are an analyst you should be flagging this as a potential advantage. On the flip side, you have to be able to adjust if your opposition has a similar advantage over you. Staying with the height advantage example, potential adjustments include playing a sweeper, or moving a midfielder or another taller defender to your full-back line to combat against your opposition’s advantage.

Trends indicate previous performance the best indication of future performance

Many coaches will lean on their performance analyst to assist in their player selection and potential substitutions. They will ask for your judgement on players because you will be expected to give an independent and fact-based evaluation rather than an opinion that may be affected by bias. This is where elite analysis can have major benefits for a team. Past behaviour is the best indicator of future behaviour, so an analyst with a good understanding of their data set will be able to assess players performances. They will notice trends and be able to estimate what each player is worth to the team based on their performances.

E.g. 1 A forward has been in form averaging 5 points a game from 7 attempts. The same forward is also giving the ball away more with poor passing and thus is costing us 3 chances a game based on his turnovers. Another forward is averaging 4 points a game along with two assists and only one turnover. By having historical data on these players we might see how forward 1 although scoring more on average is less valuable to the team. Thus in a period of a game where we are leading, and efficiency in our chances and ball retention are both paramount, we will be more likely to substitute the first forward off even if he has a better scoring record because his flaws and other players strengths are more beneficial in this situation.

E.g. 2 We might be picking a team to play an opposition who are known for two outstanding free takers but as a whole, their forwards from open play are poor particularly from the 30 metres plus range. Aware of these challenges we go about selecting a defence. We will look for players who can do several things.

1. Skilled tackler who doesn’t concede fouls

2. Defenders who can beat the forwards to contested possessions to reduce the likelihood of conceding frees in the first place

3. A sweeper who will manage a zone from sideline to sideline approximately 25-30 metres out too apply pressure on opposition forwards in their comfort zone.

4. We will allow our wing-backs to attack but ask them to defend from slightly deeper to crowd their comfort zone while asking our midfield and tracking forwards to press them between 35-50 metres away from the goal.

These two examples show how an analyst with information on their team and the opposition can integrate the data into a fact-based scenario. By designing the scenario of what we expect from the game we will be able to plan for it. The facts create a clearer picture rather than one with opinions. No scenario we plan for will appear exactly in the game, no analyst has the resources to consider every variable and the potential outcome. However by allowing facts and logic to lead your decision making you will be able to approach every decision consistently.

However, we can plan for scenarios with the management and support their decision making. These scenarios will of course remove the intangibles of players and their mentalities. No analyst should claim that they have all the facts and cannot ignore these intangibles such as leadership. If we look at the second example around selecting defenders. What if our captain, a leader who supports our young defence is prone to fouling. Are we going to remove him from the team to avoid conceding too many frees? Will his loss lead to other players in the defence being less organised? This is an inference that is much more difficult to calculate in statistics. Therefore, with this perspective, we are reminded that stats and analysis can support and even guide decision-makers but they should not be the only decision-making tool that a manager uses.